[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-openai-security-diff-scan":3,"mdc--bp9x0b-key":36,"related-org-openai-security-diff-scan":1507,"related-repo-openai-security-diff-scan":1714},{"slug":4,"name":4,"fn":5,"description":6,"org":7,"tags":11,"stars":25,"repoUrl":26,"updatedAt":27,"license":28,"forks":29,"topics":30,"repo":31,"sourceUrl":34,"mdContent":35},"security-diff-scan","perform security scans on code diffs","Use when the user asks for a security review of a pull request, commit, branch diff, working-tree patch, or other Git-backed change set.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},"openai","OpenAI","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Fopenai.png",[12,16,19,22],{"name":13,"slug":14,"type":15},"Security","security","tag",{"name":17,"slug":18,"type":15},"Static Analysis","static-analysis",{"name":20,"slug":21,"type":15},"Code Review","code-review",{"name":23,"slug":24,"type":15},"Git","git",3992,"https:\u002F\u002Fgithub.com\u002Fopenai\u002Fplugins","2026-06-30T19:00:57.102",null,465,[],{"repoUrl":26,"stars":25,"forks":29,"topics":32,"description":33},[],"OpenAI Plugins","https:\u002F\u002Fgithub.com\u002Fopenai\u002Fplugins\u002Ftree\u002FHEAD\u002Fplugins\u002Fcodex-security\u002Fskills\u002Fsecurity-diff-scan","---\nname: security-diff-scan\ndescription: \"Use when the user asks for a security review of a pull request, commit, branch diff, working-tree patch, or other Git-backed change set.\"\n---\n\n# Security Diff Scan\n\nUsed when a user wants to review a Git-backed change set for security regressions. Keep the scan phases separate and produce the final markdown report.\n\n## Setup Workspace Routing\n\nWhen this skill is the active top-level workflow, use the setup workspace only when the host context explicitly says it is running inside the Codex desktop app and both required setup continuation tools are available. Tool availability alone does not identify the app host. Otherwise, including Codex CLI interactive and headless runs, use the prompt-only terminal\u002Fchat workflow: do not call Codex Security app setup tools, ask the user to press Start scan, or wait for an app-generated `scanId`.\n\nTreat goal creation as scan execution, not setup. In the app setup path, do not create or adopt scan goals before the user presses Start scan, the authoritative scan context has been loaded from a `status: \"started\"` wait result or a direct continuation with a `scanId`, and the capability preflight has returned `ready`.\n\nFor an app continuation that already includes a `scanId` and optional `handoffClaimToken`, do not open another workspace: call `get_codex_security_scan_context` with the `scanId`, pass its `handoffClaimToken` when present, route elsewhere only if its validated mode differs, and use its target, `diffTarget`, optional `userContext`, and `scanDir`.\n\nOtherwise, in a host that renders MCP Apps and exposes the Codex Security setup continuation tools:\n\n1. Resolve setup arguments directly from the user's initial prompt and known thread context: checked-out Git repository `targetPath`, `mode: \"diff\"`, `scope: \".\"`, user-supplied security focus as `userContext`, and `diffTarget` only when the prompt unambiguously identifies uncommitted changes against current `HEAD`, one commit, or a locally resolved PR, branch comparison, or revision range.\n2. Perform only the minimal path or revision resolution needed to construct those arguments. Do not run capability preflight, inspect the repository beyond that minimal resolution, threat model, discover findings, or create workers before setup opens.\n3. Immediately call `open_codex_security_workspace` with the resolved arguments. Do not search for or substitute a separate scan command.\n4. Immediately call `await_codex_security_scan_start` with the `sessionId` from the workspace returned by `open_codex_security_workspace`. A returned workspace with `setup.submitted=false` is the expected wait state. Keep the tool call pending while waiting for the user to review setup and press Start scan; do not create or adopt a scan goal, run preflight, or pivot to terminal\u002Fchat fallback while waiting.\n5. If the wait returns `status: \"started\"`, require its `scanId`, call `get_codex_security_scan_context` with that `scanId`, and pass its `handoffClaimToken` when present. Then run the preflight in `..\u002F..\u002Freferences\u002Fconfig-preflight.md` for the selected target and `security_diff_scan` profile before goal setup, threat modeling, or other substantive scan work.\n6. If the wait returns `status: \"already_delivered\"`, end the current turn without loading scan context or starting scan work. Another continuation already owns the scan.\n7. If the wait returns `status: \"timed_out\"`, end the current turn and tell the user to finish setup and use **Continue in Codex** after pressing Start scan. Do not run preflight, create or adopt a scan goal, open another workspace, or pivot to terminal\u002Fchat fallback.\n8. Continue after a `ready` result, explaining material warn or suggest limitations. If preflight is `blocked` or `incomplete` with actionable remediation, present the exact reasons and config delta, ask whether to apply the remediation, and stop for the user's answer before creating or adopting a scan goal or calling `fail_codex_security_scan`. Do not fail automatically for declined or unavailable remediation, helper errors, or a non-ready rerun. Preserve the running scan and retry or hand off while recovery may still be possible. If the user declines required remediation, ask whether to cancel or leave the scan running for a later retry. Call `fail_codex_security_scan` with the exact reason only after documented recovery is exhausted and the blocker is confirmed unrecoverable, or when the user explicitly cancels.\n\nBefore opening setup, use the existing terminal\u002Fchat preflight and scan workflow for local changes against another requested base because the setup app cannot represent that working-tree diff target. Codex CLI, including interactive and headless runs, and hosts without the required app capabilities use the same prompt-only fallback. Do not call `open_codex_security_workspace` or `await_codex_security_scan_start` on this path. Once `open_codex_security_workspace` succeeds in an MCP Apps-capable host, remain on the app path: immediately call `await_codex_security_scan_start`; a `status: \"timed_out\"` result means end the turn and point the user to **Continue in Codex**, while `status: \"already_delivered\"` means stop because another continuation owns the scan. Do not start a terminal\u002Fchat fallback for either result.\n\n## Capability Preflight\n\nRead `..\u002F..\u002Freferences\u002Fconfig-preflight.md` and dispatch and await the preflight execution described there with the `security_diff_scan` capability profile before substantive scan work, including after an app wait or direct continuation has produced a `scanId` and loaded its authoritative scan context. Follow the returned block\u002Fwarn\u002Fsuggest results. For an app-generated scan, ask before applying actionable remediation and wait without creating a scan goal or calling `fail_codex_security_scan`. Do not fail automatically for declined or unavailable remediation, helper errors, or a non-ready rerun; preserve the running scan and retry or hand off while recovery may still be possible. Call `fail_codex_security_scan` only after documented recovery is exhausted and the blocker is confirmed unrecoverable, or when the user explicitly cancels. Do not treat a config value that differs from a suggested patch as a warning unless the capability requirement itself is unmet.\n\n## Phase Sequence\n\nKeep these phases distinct and run them in linear order:\n\n1. `$threat-model`\n2. `$finding-discovery`\n3. `$validation`\n4. `$attack-path-analysis`\n5. Generate final output\n\nTreat this skill as the top-level orchestrator for the four skills plus the final report assembly step. Do not collapse the phases together.\n\nFor each phase:\n1. Read that phase's skill.\n2. Load only the inputs required for that phase.\n3. Complete that phase's workflow and checklist.\n4. Only then read the next phase's skill.\n\nDo not read ahead into later-phase skills until the current phase has completed.\nDo not amortize effort across phases: complete each phase to the full depth expected by that phase before moving on.\nTreat explicit invocation of this exhaustive diff-scan workflow as the user's authorization to use the subagents required by the workflow. If subagents are unavailable in the current environment, explain the limitation instead of claiming exhaustive diff coverage.\n\n## Goal Setup\n\nAfter the app wait or direct continuation has provided a `scanId`, the authoritative scan context has been loaded, and the `security_diff_scan` capability preflight has returned `ready`, or after the same preflight is `ready` in Codex CLI or terminal\u002Fchat hosts without the setup app, create a Codex goal for the scan if the runtime exposes goal tools and no active goal already covers this scan. The objective should state that the scan must not stop until the resolved diff-scoped files have been covered and the required coverage artifacts prove that closure.\n\nUse objective wording shaped like:\n\n`Run the Codex Security diff scan for \u003Cresolved target>; do not stop until every diff-scoped file\u002Fworklist row has a completion receipt or explicit deferred closure, every candidate has required ledger receipts, and the final report is written.`\n\nIf a compatible active goal already exists, continue under it instead of creating a duplicate. If goal tools are unavailable, state the same coverage objective in the first visible scan update and continue.\n\nDo not mark the goal complete until:\n\n- every `deep_review_input.jsonl` row has a completion receipt in `work_ledger.jsonl`, or an explicit `deferred`, `not_applicable`, or `suppressed` closure with exact reason\n- every candidate that reached discovery has the required discovery, validation, and attack-path ledger receipts, or an explicit deferred reason for the missing proof\n- the final markdown report has been written to the resolved scan path\n\n## Artifact Resolution\n\nThe path references in this skill are the default locations for this phase.\nIf the user explicitly provides a different path for a required input or output, use the user-provided path instead of the corresponding default path referenced in this skill.\nIf a required input is still missing, stop and ask the user for it before continuing.\nUse the shared scan artifact path conventions in `..\u002F..\u002Freferences\u002Fscan-artifacts.md`.\n\n## Execution Plan\n\nStart this plan only after `Setup Workspace Routing` has either loaded the app-generated scan context with a `scanId`, or determined that the host is using the non-app terminal\u002Fchat workflow, and the `security_diff_scan` capability preflight has returned `ready`.\n\nFollow this plan in order. Do not skip ahead to a later phase until the current phase has produced its intended output.\n\n1. Resolve the Git-backed scan target, `repo_name`, `security_scans_dir`, `scan_id`, `scan_dir`, and `artifacts_dir` using `..\u002F..\u002Freferences\u002Fscan-artifacts.md`.\n2. Create or adopt the scan goal described in `Goal Setup` for that active scan context.\n3. Read `..\u002F..\u002Freferences\u002Fsecurity-guidance.md`, compile the repository's policy to `\u003Ccontext_dir>\u002Fsecurity_guidance.md`, and read it before threat modeling or inspecting source code.\n4. Run `$threat-model` first.\n  - Copy the repository-scoped threat model to the per-scan threat model path without alteration for auditability.\n  - Treat the per-scan threat model path as the source of truth threat model for later phases.\n5. Run `$finding-discovery` as the second step, against the resolved diff and using the per-scan threat model as context.\n  - If discovery produces no technically plausible candidates, stop there, skip validation and attack-path analysis, complete the canonical JSON contract, and finalize the scan.\n6. Run `$validation` as the third step, for each candidate that came out of discovery.\n  - Pass the resolved diff scope, discovery notes, and candidate inventory to validation. Validation should preserve or suppress the provided instances; it should not independently broaden the review into a repository-wide scan.\n  - Each candidate finding's `findings\u002F\u003Ccandidate_id>\u002Fcandidate_ledger.jsonl` is part of the validation input. Every candidate finding that came out of discovery must have a discovery receipt before validation starts and a validation receipt before the scan can proceed to final reporting.\n7. Run `$attack-path-analysis` as the fourth step, for findings that still need reportability, attack-path, and severity analysis after validation.\n  - Each candidate finding's `findings\u002F\u003Ccandidate_id>\u002Fcandidate_ledger.jsonl` is part of the attack-path input. Every candidate finding that reaches attack-path analysis must have an attack-path receipt before final reporting, even when the final decision is `ignore`, suppressed, or deferred.\n8. Assemble the complete canonical JSON contract last using `..\u002F..\u002Freferences\u002Ffinal-report.md`; do not author `report.md`.\n  - Populate the optional structured details in `..\u002F..\u002Freferences\u002Ffinding-detail-fields.md` from the same validated evidence used in the generated report.\n  - For every reportable finding, run `$vulnerability-writeup` with exactly one dedicated write-up sub-agent. Give it only that finding, its validation and attack-path evidence, relevant source paths and revision, PoC inputs, and the target output directory.\n  - Write the derived report to `findings\u002F\u003Cslug>\u002F\u003Cslug>.md` with supporting PoC files under `findings\u002F\u003Cslug>\u002Fpoc\u002F`. Verify the report is a regular file, then set that finding's `writeup.reportPath` to the matching safe relative path. Do not add the derived report to the sealed artifact list.\n  - After every write-up is ready, run `$propose-security-hardening` once over the complete finding collection, detailed write-ups, threat model, coverage, and relevant source. Write its portfolio to `hardening\u002Fhardening.md`, its structured analysis to `hardening\u002Fhardening.json`, and any proposals and diagrams below `hardening\u002F`. Verify `hardening\u002Fhardening.md` is a regular file, then set `scan.hardening.portfolioPath` to the fixed relative path `hardening\u002Fhardening.md`. Do not add these derived files to the sealed artifact list. Skip this step and omit `scan.hardening` when there are no reportable findings.\n  - Complete the scan once, after all write-ups, hardening guidance, and canonical JSON are ready, so finalization projects the validated JSON and derived-document links into `report.md`. In the terminal\u002Fchat workflow without `complete_codex_security_scan`, run `python \u003Cplugin_dir>\u002Fscripts\u002Ffinalize_scan_contract.py --scan-dir \u003Cscan_dir> --source-root \u003Crepo_root>` directly.\n\n## Phase Scope\n\n- Phase 1 (threat model generation) is repository-scope by default, unless the user explicitly asks for narrower scope or provides an authoritative threat model or sufficiently repository-specific security scan guidance such as `AGENTS.md`.\n- Phase 2 onward (finding discovery, validation, attack path analysis) are diff-focused and should follow the changed code and its supporting files.\n\nTreat this asymmetry as intentional:\n\n- use the diff to locate the scan target for later phases\n- do not let the diff bias Phase 1 threat model generation, if applicable\n- do not let the touched subsystem become the repository threat model unless the user explicitly asks for that narrower scope\n\n## Scan Target\n\nResolve the exact Git-backed diff before starting:\n\n- PR: compare base branch against current `HEAD`\n- commit: scan the target commit against its parent or requested baseline\n- branch diff: scan the requested merge-base to head range\n- local patch: scan staged and unstaged working-tree changes against the requested base\n\n## Diff-Scoped Discovery\n\nUse `..\u002Fsecurity-scan\u002Freferences\u002Fscan-artifacts-and-ledger.md` for the shared scoped file-review, candidate-ledger, subagent, and dedupe rules.\n\nDiff scans should:\n\n- generate `rank_input.jsonl` deterministically from changed source-like files with `\u003Cpython_command> \u003Cplugin_dir>\u002Fscripts\u002Fgenerate_rank_input.py make-diff-rank-input --repo \u003Crepo_root> --base \u003Cbase> --mode revisions --head \u003Chead> --out \u003Cdiscovery_dir>\u002Frank_input.jsonl` for PR, commit, and branch diffs, or `\u003Cpython_command> \u003Cplugin_dir>\u002Fscripts\u002Fgenerate_rank_input.py make-diff-rank-input --repo \u003Crepo_root> --base \u003Cbase> --mode local-patch --out \u003Cdiscovery_dir>\u002Frank_input.jsonl` for a local patch\n- copy every diff row into `deep_review_input.jsonl` with `\u003Cpython_command> \u003Cplugin_dir>\u002Fscripts\u002Fgenerate_rank_input.py copy-deep-review-input --rank-input \u003Cdiscovery_dir>\u002Frank_input.jsonl --out \u003Cdiscovery_dir>\u002Fdeep_review_input.jsonl`\n- deep-review every file in `deep_review_input.jsonl`\n- add directly supporting files only when repository evidence shows they are needed to understand the changed security behavior\n- stay anchored to the changed code and directly supporting files rather than broadening into unrelated repository-wide enumeration\n\n## Diff-Scoped Sibling Coverage\n\nFor PR, commit, branch, and local-patch scans, stay diff-focused but preserve repeated vulnerable instances that are created or affected by the same changed pattern.\n\nDiff scans should:\n\n- start from the changed files and the supporting files needed to understand the changed behavior\n- expand from a changed route, handler, shared helper, guard, template pattern, query builder, serializer\u002Fdeserializer, filesystem\u002Fnetwork sink, config block, or wrapper to sibling instances that the diff also changes, newly reaches, or affects through the same modified shared dependency\n- when the diff adds, removes, or reshapes a guard around an existing parser, deserializer, expression evaluator, filesystem\u002Fpath helper, archive utility, or auth\u002Fauthz helper, use the adjacent pre-existing sink\u002Fcontrol as supporting context for the changed behavior; keep the candidate anchored to the changed guard or newly exposed path unless the user explicitly asks for wider instance expansion\n- when a changed wrapper, guard, or API delegates to a shared parser\u002Fdeserializer\u002Fpath\u002Farchive\u002Fauth helper, keep both the wrapper call site and the underlying shared sink\u002Fcontrol line addressable; do not replace the root sink\u002Fcontrol evidence with wrapper-only evidence\n- carry each vulnerable sibling instance through discovery and validation with its own affected location, source, closest control, sink, impact, and suppression evidence\n- use unchanged siblings as context and negative controls, but report them only when the diff makes them newly vulnerable or changes the shared control or sink they depend on\n- stop when the diff-linked pattern family is exhausted, rather than broadening into repository-wide enumeration\n\nThis keeps diff scans precise while avoiding the common failure mode where one representative route or sink hides additional vulnerable siblings introduced by the same patch.\n\n## Final Output\n\nPopulate all final report semantics in the canonical manifest, findings, and coverage JSON using `..\u002F..\u002Freferences\u002Ffinal-report.md`. Generate one detailed `vulnerability-writeup` for every reportable finding, then run `propose-security-hardening` once over the complete collection and record the safe derived-document paths. Complete the scan once after both stages; finalization owns `report.md` generation. Emit Codex app review directives from the completed canonical findings. Commit scans use this same final-output contract because they are a diff-scan target type.\n\n## Hard Rules\n\nRead `..\u002F..\u002Freferences\u002Fshared-hard-rules.md` before applying scan-mode-specific hard rules.\n\n- After any app setup handoff has provided a `scanId`, or in the non-app terminal\u002Fchat workflow, create or adopt the scan goal only after the capability preflight has returned `ready`, and before substantive scan work. Do not complete it until the resolved diff-scoped files\u002Fworklist rows, candidate ledgers, and final report meet the `Goal Setup` closure criteria.\n- Do not claim diff coverage until every `deep_review_input.jsonl` row has a completion receipt in `work_ledger.jsonl`.\n",{"data":37,"body":38},{"name":4,"description":6},{"type":39,"children":40},"root",[41,49,55,62,76,103,168,173,410,462,468,508,514,519,563,568,573,596,601,607,640,645,654,659,664,722,728,740,746,777,782,876,889,904,912,927,948,963,985,1009,1162,1168,1188,1193,1211,1217,1222,1250,1256,1269,1274,1344,1350,1355,1359,1397,1402,1408,1443,1449,1461],{"type":42,"tag":43,"props":44,"children":45},"element","h1",{"id":4},[46],{"type":47,"value":48},"text","Security Diff Scan",{"type":42,"tag":50,"props":51,"children":52},"p",{},[53],{"type":47,"value":54},"Used when a user wants to review a Git-backed change set for security regressions. Keep the scan phases separate and produce the final markdown report.",{"type":42,"tag":56,"props":57,"children":59},"h2",{"id":58},"setup-workspace-routing",[60],{"type":47,"value":61},"Setup Workspace Routing",{"type":42,"tag":50,"props":63,"children":64},{},[65,67,74],{"type":47,"value":66},"When this skill is the active top-level workflow, use the setup workspace only when the host context explicitly says it is running inside the Codex desktop app and both required setup continuation tools are available. Tool availability alone does not identify the app host. Otherwise, including Codex CLI interactive and headless runs, use the prompt-only terminal\u002Fchat workflow: do not call Codex Security app setup tools, ask the user to press Start scan, or wait for an app-generated ",{"type":42,"tag":68,"props":69,"children":71},"code",{"className":70},[],[72],{"type":47,"value":73},"scanId",{"type":47,"value":75},".",{"type":42,"tag":50,"props":77,"children":78},{},[79,81,87,89,94,96,102],{"type":47,"value":80},"Treat goal creation as scan execution, not setup. In the app setup path, do not create or adopt scan goals before the user presses Start scan, the authoritative scan context has been loaded from a ",{"type":42,"tag":68,"props":82,"children":84},{"className":83},[],[85],{"type":47,"value":86},"status: \"started\"",{"type":47,"value":88}," wait result or a direct continuation with a ",{"type":42,"tag":68,"props":90,"children":92},{"className":91},[],[93],{"type":47,"value":73},{"type":47,"value":95},", and the capability preflight has returned ",{"type":42,"tag":68,"props":97,"children":99},{"className":98},[],[100],{"type":47,"value":101},"ready",{"type":47,"value":75},{"type":42,"tag":50,"props":104,"children":105},{},[106,108,113,115,121,123,129,131,136,138,143,145,151,153,159,161,167],{"type":47,"value":107},"For an app continuation that already includes a ",{"type":42,"tag":68,"props":109,"children":111},{"className":110},[],[112],{"type":47,"value":73},{"type":47,"value":114}," and optional ",{"type":42,"tag":68,"props":116,"children":118},{"className":117},[],[119],{"type":47,"value":120},"handoffClaimToken",{"type":47,"value":122},", do not open another workspace: call ",{"type":42,"tag":68,"props":124,"children":126},{"className":125},[],[127],{"type":47,"value":128},"get_codex_security_scan_context",{"type":47,"value":130}," with the ",{"type":42,"tag":68,"props":132,"children":134},{"className":133},[],[135],{"type":47,"value":73},{"type":47,"value":137},", pass its ",{"type":42,"tag":68,"props":139,"children":141},{"className":140},[],[142],{"type":47,"value":120},{"type":47,"value":144}," when present, route elsewhere only if its validated mode differs, and use its target, ",{"type":42,"tag":68,"props":146,"children":148},{"className":147},[],[149],{"type":47,"value":150},"diffTarget",{"type":47,"value":152},", optional ",{"type":42,"tag":68,"props":154,"children":156},{"className":155},[],[157],{"type":47,"value":158},"userContext",{"type":47,"value":160},", and ",{"type":42,"tag":68,"props":162,"children":164},{"className":163},[],[165],{"type":47,"value":166},"scanDir",{"type":47,"value":75},{"type":42,"tag":50,"props":169,"children":170},{},[171],{"type":47,"value":172},"Otherwise, in a host that renders MCP Apps and exposes the Codex Security setup continuation tools:",{"type":42,"tag":174,"props":175,"children":176},"ol",{},[177,227,232,245,279,335,347,367],{"type":42,"tag":178,"props":179,"children":180},"li",{},[181,183,189,191,197,198,204,206,211,212,217,219,225],{"type":47,"value":182},"Resolve setup arguments directly from the user's initial prompt and known thread context: checked-out Git repository ",{"type":42,"tag":68,"props":184,"children":186},{"className":185},[],[187],{"type":47,"value":188},"targetPath",{"type":47,"value":190},", ",{"type":42,"tag":68,"props":192,"children":194},{"className":193},[],[195],{"type":47,"value":196},"mode: \"diff\"",{"type":47,"value":190},{"type":42,"tag":68,"props":199,"children":201},{"className":200},[],[202],{"type":47,"value":203},"scope: \".\"",{"type":47,"value":205},", user-supplied security focus as ",{"type":42,"tag":68,"props":207,"children":209},{"className":208},[],[210],{"type":47,"value":158},{"type":47,"value":160},{"type":42,"tag":68,"props":213,"children":215},{"className":214},[],[216],{"type":47,"value":150},{"type":47,"value":218}," only when the prompt unambiguously identifies uncommitted changes against current ",{"type":42,"tag":68,"props":220,"children":222},{"className":221},[],[223],{"type":47,"value":224},"HEAD",{"type":47,"value":226},", one commit, or a locally resolved PR, branch comparison, or revision range.",{"type":42,"tag":178,"props":228,"children":229},{},[230],{"type":47,"value":231},"Perform only the minimal path or revision resolution needed to construct those arguments. Do not run capability preflight, inspect the repository beyond that minimal resolution, threat model, discover findings, or create workers before setup opens.",{"type":42,"tag":178,"props":233,"children":234},{},[235,237,243],{"type":47,"value":236},"Immediately call ",{"type":42,"tag":68,"props":238,"children":240},{"className":239},[],[241],{"type":47,"value":242},"open_codex_security_workspace",{"type":47,"value":244}," with the resolved arguments. Do not search for or substitute a separate scan command.",{"type":42,"tag":178,"props":246,"children":247},{},[248,249,255,256,262,264,269,271,277],{"type":47,"value":236},{"type":42,"tag":68,"props":250,"children":252},{"className":251},[],[253],{"type":47,"value":254},"await_codex_security_scan_start",{"type":47,"value":130},{"type":42,"tag":68,"props":257,"children":259},{"className":258},[],[260],{"type":47,"value":261},"sessionId",{"type":47,"value":263}," from the workspace returned by ",{"type":42,"tag":68,"props":265,"children":267},{"className":266},[],[268],{"type":47,"value":242},{"type":47,"value":270},". A returned workspace with ",{"type":42,"tag":68,"props":272,"children":274},{"className":273},[],[275],{"type":47,"value":276},"setup.submitted=false",{"type":47,"value":278}," is the expected wait state. Keep the tool call pending while waiting for the user to review setup and press Start scan; do not create or adopt a scan goal, run preflight, or pivot to terminal\u002Fchat fallback while waiting.",{"type":42,"tag":178,"props":280,"children":281},{},[282,284,289,291,296,298,303,305,310,312,317,319,325,327,333],{"type":47,"value":283},"If the wait returns ",{"type":42,"tag":68,"props":285,"children":287},{"className":286},[],[288],{"type":47,"value":86},{"type":47,"value":290},", require its ",{"type":42,"tag":68,"props":292,"children":294},{"className":293},[],[295],{"type":47,"value":73},{"type":47,"value":297},", call ",{"type":42,"tag":68,"props":299,"children":301},{"className":300},[],[302],{"type":47,"value":128},{"type":47,"value":304}," with that ",{"type":42,"tag":68,"props":306,"children":308},{"className":307},[],[309],{"type":47,"value":73},{"type":47,"value":311},", and pass its ",{"type":42,"tag":68,"props":313,"children":315},{"className":314},[],[316],{"type":47,"value":120},{"type":47,"value":318}," when present. Then run the preflight in ",{"type":42,"tag":68,"props":320,"children":322},{"className":321},[],[323],{"type":47,"value":324},"..\u002F..\u002Freferences\u002Fconfig-preflight.md",{"type":47,"value":326}," for the selected target and ",{"type":42,"tag":68,"props":328,"children":330},{"className":329},[],[331],{"type":47,"value":332},"security_diff_scan",{"type":47,"value":334}," profile before goal setup, threat modeling, or other substantive scan work.",{"type":42,"tag":178,"props":336,"children":337},{},[338,339,345],{"type":47,"value":283},{"type":42,"tag":68,"props":340,"children":342},{"className":341},[],[343],{"type":47,"value":344},"status: \"already_delivered\"",{"type":47,"value":346},", end the current turn without loading scan context or starting scan work. Another continuation already owns the scan.",{"type":42,"tag":178,"props":348,"children":349},{},[350,351,357,359,365],{"type":47,"value":283},{"type":42,"tag":68,"props":352,"children":354},{"className":353},[],[355],{"type":47,"value":356},"status: \"timed_out\"",{"type":47,"value":358},", end the current turn and tell the user to finish setup and use ",{"type":42,"tag":360,"props":361,"children":362},"strong",{},[363],{"type":47,"value":364},"Continue in Codex",{"type":47,"value":366}," after pressing Start scan. Do not run preflight, create or adopt a scan goal, open another workspace, or pivot to terminal\u002Fchat fallback.",{"type":42,"tag":178,"props":368,"children":369},{},[370,372,377,379,385,387,393,395,401,403,408],{"type":47,"value":371},"Continue after a ",{"type":42,"tag":68,"props":373,"children":375},{"className":374},[],[376],{"type":47,"value":101},{"type":47,"value":378}," result, explaining material warn or suggest limitations. If preflight is ",{"type":42,"tag":68,"props":380,"children":382},{"className":381},[],[383],{"type":47,"value":384},"blocked",{"type":47,"value":386}," or ",{"type":42,"tag":68,"props":388,"children":390},{"className":389},[],[391],{"type":47,"value":392},"incomplete",{"type":47,"value":394}," with actionable remediation, present the exact reasons and config delta, ask whether to apply the remediation, and stop for the user's answer before creating or adopting a scan goal or calling ",{"type":42,"tag":68,"props":396,"children":398},{"className":397},[],[399],{"type":47,"value":400},"fail_codex_security_scan",{"type":47,"value":402},". Do not fail automatically for declined or unavailable remediation, helper errors, or a non-ready rerun. Preserve the running scan and retry or hand off while recovery may still be possible. If the user declines required remediation, ask whether to cancel or leave the scan running for a later retry. Call ",{"type":42,"tag":68,"props":404,"children":406},{"className":405},[],[407],{"type":47,"value":400},{"type":47,"value":409}," with the exact reason only after documented recovery is exhausted and the blocker is confirmed unrecoverable, or when the user explicitly cancels.",{"type":42,"tag":50,"props":411,"children":412},{},[413,415,420,421,426,428,433,435,440,442,447,449,453,455,460],{"type":47,"value":414},"Before opening setup, use the existing terminal\u002Fchat preflight and scan workflow for local changes against another requested base because the setup app cannot represent that working-tree diff target. Codex CLI, including interactive and headless runs, and hosts without the required app capabilities use the same prompt-only fallback. Do not call ",{"type":42,"tag":68,"props":416,"children":418},{"className":417},[],[419],{"type":47,"value":242},{"type":47,"value":386},{"type":42,"tag":68,"props":422,"children":424},{"className":423},[],[425],{"type":47,"value":254},{"type":47,"value":427}," on this path. Once ",{"type":42,"tag":68,"props":429,"children":431},{"className":430},[],[432],{"type":47,"value":242},{"type":47,"value":434}," succeeds in an MCP Apps-capable host, remain on the app path: immediately call ",{"type":42,"tag":68,"props":436,"children":438},{"className":437},[],[439],{"type":47,"value":254},{"type":47,"value":441},"; a ",{"type":42,"tag":68,"props":443,"children":445},{"className":444},[],[446],{"type":47,"value":356},{"type":47,"value":448}," result means end the turn and point the user to ",{"type":42,"tag":360,"props":450,"children":451},{},[452],{"type":47,"value":364},{"type":47,"value":454},", while ",{"type":42,"tag":68,"props":456,"children":458},{"className":457},[],[459],{"type":47,"value":344},{"type":47,"value":461}," means stop because another continuation owns the scan. Do not start a terminal\u002Fchat fallback for either result.",{"type":42,"tag":56,"props":463,"children":465},{"id":464},"capability-preflight",[466],{"type":47,"value":467},"Capability Preflight",{"type":42,"tag":50,"props":469,"children":470},{},[471,473,478,480,485,487,492,494,499,501,506],{"type":47,"value":472},"Read ",{"type":42,"tag":68,"props":474,"children":476},{"className":475},[],[477],{"type":47,"value":324},{"type":47,"value":479}," and dispatch and await the preflight execution described there with the ",{"type":42,"tag":68,"props":481,"children":483},{"className":482},[],[484],{"type":47,"value":332},{"type":47,"value":486}," capability profile before substantive scan work, including after an app wait or direct continuation has produced a ",{"type":42,"tag":68,"props":488,"children":490},{"className":489},[],[491],{"type":47,"value":73},{"type":47,"value":493}," and loaded its authoritative scan context. Follow the returned block\u002Fwarn\u002Fsuggest results. For an app-generated scan, ask before applying actionable remediation and wait without creating a scan goal or calling ",{"type":42,"tag":68,"props":495,"children":497},{"className":496},[],[498],{"type":47,"value":400},{"type":47,"value":500},". Do not fail automatically for declined or unavailable remediation, helper errors, or a non-ready rerun; preserve the running scan and retry or hand off while recovery may still be possible. Call ",{"type":42,"tag":68,"props":502,"children":504},{"className":503},[],[505],{"type":47,"value":400},{"type":47,"value":507}," only after documented recovery is exhausted and the blocker is confirmed unrecoverable, or when the user explicitly cancels. Do not treat a config value that differs from a suggested patch as a warning unless the capability requirement itself is unmet.",{"type":42,"tag":56,"props":509,"children":511},{"id":510},"phase-sequence",[512],{"type":47,"value":513},"Phase Sequence",{"type":42,"tag":50,"props":515,"children":516},{},[517],{"type":47,"value":518},"Keep these phases distinct and run them in linear order:",{"type":42,"tag":174,"props":520,"children":521},{},[522,531,540,549,558],{"type":42,"tag":178,"props":523,"children":524},{},[525],{"type":42,"tag":68,"props":526,"children":528},{"className":527},[],[529],{"type":47,"value":530},"$threat-model",{"type":42,"tag":178,"props":532,"children":533},{},[534],{"type":42,"tag":68,"props":535,"children":537},{"className":536},[],[538],{"type":47,"value":539},"$finding-discovery",{"type":42,"tag":178,"props":541,"children":542},{},[543],{"type":42,"tag":68,"props":544,"children":546},{"className":545},[],[547],{"type":47,"value":548},"$validation",{"type":42,"tag":178,"props":550,"children":551},{},[552],{"type":42,"tag":68,"props":553,"children":555},{"className":554},[],[556],{"type":47,"value":557},"$attack-path-analysis",{"type":42,"tag":178,"props":559,"children":560},{},[561],{"type":47,"value":562},"Generate final output",{"type":42,"tag":50,"props":564,"children":565},{},[566],{"type":47,"value":567},"Treat this skill as the top-level orchestrator for the four skills plus the final report assembly step. Do not collapse the phases together.",{"type":42,"tag":50,"props":569,"children":570},{},[571],{"type":47,"value":572},"For each phase:",{"type":42,"tag":174,"props":574,"children":575},{},[576,581,586,591],{"type":42,"tag":178,"props":577,"children":578},{},[579],{"type":47,"value":580},"Read that phase's skill.",{"type":42,"tag":178,"props":582,"children":583},{},[584],{"type":47,"value":585},"Load only the inputs required for that phase.",{"type":42,"tag":178,"props":587,"children":588},{},[589],{"type":47,"value":590},"Complete that phase's workflow and checklist.",{"type":42,"tag":178,"props":592,"children":593},{},[594],{"type":47,"value":595},"Only then read the next phase's skill.",{"type":42,"tag":50,"props":597,"children":598},{},[599],{"type":47,"value":600},"Do not read ahead into later-phase skills until the current phase has completed.\nDo not amortize effort across phases: complete each phase to the full depth expected by that phase before moving on.\nTreat explicit invocation of this exhaustive diff-scan workflow as the user's authorization to use the subagents required by the workflow. If subagents are unavailable in the current environment, explain the limitation instead of claiming exhaustive diff coverage.",{"type":42,"tag":56,"props":602,"children":604},{"id":603},"goal-setup",[605],{"type":47,"value":606},"Goal Setup",{"type":42,"tag":50,"props":608,"children":609},{},[610,612,617,619,624,626,631,633,638],{"type":47,"value":611},"After the app wait or direct continuation has provided a ",{"type":42,"tag":68,"props":613,"children":615},{"className":614},[],[616],{"type":47,"value":73},{"type":47,"value":618},", the authoritative scan context has been loaded, and the ",{"type":42,"tag":68,"props":620,"children":622},{"className":621},[],[623],{"type":47,"value":332},{"type":47,"value":625}," capability preflight has returned ",{"type":42,"tag":68,"props":627,"children":629},{"className":628},[],[630],{"type":47,"value":101},{"type":47,"value":632},", or after the same preflight is ",{"type":42,"tag":68,"props":634,"children":636},{"className":635},[],[637],{"type":47,"value":101},{"type":47,"value":639}," in Codex CLI or terminal\u002Fchat hosts without the setup app, create a Codex goal for the scan if the runtime exposes goal tools and no active goal already covers this scan. The objective should state that the scan must not stop until the resolved diff-scoped files have been covered and the required coverage artifacts prove that closure.",{"type":42,"tag":50,"props":641,"children":642},{},[643],{"type":47,"value":644},"Use objective wording shaped like:",{"type":42,"tag":50,"props":646,"children":647},{},[648],{"type":42,"tag":68,"props":649,"children":651},{"className":650},[],[652],{"type":47,"value":653},"Run the Codex Security diff scan for \u003Cresolved target>; do not stop until every diff-scoped file\u002Fworklist row has a completion receipt or explicit deferred closure, every candidate has required ledger receipts, and the final report is written.",{"type":42,"tag":50,"props":655,"children":656},{},[657],{"type":47,"value":658},"If a compatible active goal already exists, continue under it instead of creating a duplicate. If goal tools are unavailable, state the same coverage objective in the first visible scan update and continue.",{"type":42,"tag":50,"props":660,"children":661},{},[662],{"type":47,"value":663},"Do not mark the goal complete until:",{"type":42,"tag":665,"props":666,"children":667},"ul",{},[668,712,717],{"type":42,"tag":178,"props":669,"children":670},{},[671,673,679,681,687,689,695,696,702,704,710],{"type":47,"value":672},"every ",{"type":42,"tag":68,"props":674,"children":676},{"className":675},[],[677],{"type":47,"value":678},"deep_review_input.jsonl",{"type":47,"value":680}," row has a completion receipt in ",{"type":42,"tag":68,"props":682,"children":684},{"className":683},[],[685],{"type":47,"value":686},"work_ledger.jsonl",{"type":47,"value":688},", or an explicit ",{"type":42,"tag":68,"props":690,"children":692},{"className":691},[],[693],{"type":47,"value":694},"deferred",{"type":47,"value":190},{"type":42,"tag":68,"props":697,"children":699},{"className":698},[],[700],{"type":47,"value":701},"not_applicable",{"type":47,"value":703},", or ",{"type":42,"tag":68,"props":705,"children":707},{"className":706},[],[708],{"type":47,"value":709},"suppressed",{"type":47,"value":711}," closure with exact reason",{"type":42,"tag":178,"props":713,"children":714},{},[715],{"type":47,"value":716},"every candidate that reached discovery has the required discovery, validation, and attack-path ledger receipts, or an explicit deferred reason for the missing proof",{"type":42,"tag":178,"props":718,"children":719},{},[720],{"type":47,"value":721},"the final markdown report has been written to the resolved scan path",{"type":42,"tag":56,"props":723,"children":725},{"id":724},"artifact-resolution",[726],{"type":47,"value":727},"Artifact Resolution",{"type":42,"tag":50,"props":729,"children":730},{},[731,733,739],{"type":47,"value":732},"The path references in this skill are the default locations for this phase.\nIf the user explicitly provides a different path for a required input or output, use the user-provided path instead of the corresponding default path referenced in this skill.\nIf a required input is still missing, stop and ask the user for it before continuing.\nUse the shared scan artifact path conventions in ",{"type":42,"tag":68,"props":734,"children":736},{"className":735},[],[737],{"type":47,"value":738},"..\u002F..\u002Freferences\u002Fscan-artifacts.md",{"type":47,"value":75},{"type":42,"tag":56,"props":741,"children":743},{"id":742},"execution-plan",[744],{"type":47,"value":745},"Execution Plan",{"type":42,"tag":50,"props":747,"children":748},{},[749,751,756,758,763,765,770,771,776],{"type":47,"value":750},"Start this plan only after ",{"type":42,"tag":68,"props":752,"children":754},{"className":753},[],[755],{"type":47,"value":61},{"type":47,"value":757}," has either loaded the app-generated scan context with a ",{"type":42,"tag":68,"props":759,"children":761},{"className":760},[],[762],{"type":47,"value":73},{"type":47,"value":764},", or determined that the host is using the non-app terminal\u002Fchat workflow, and the ",{"type":42,"tag":68,"props":766,"children":768},{"className":767},[],[769],{"type":47,"value":332},{"type":47,"value":625},{"type":42,"tag":68,"props":772,"children":774},{"className":773},[],[775],{"type":47,"value":101},{"type":47,"value":75},{"type":42,"tag":50,"props":778,"children":779},{},[780],{"type":47,"value":781},"Follow this plan in order. Do not skip ahead to a later phase until the current phase has produced its intended output.",{"type":42,"tag":174,"props":783,"children":784},{},[785,832,844,864],{"type":42,"tag":178,"props":786,"children":787},{},[788,790,796,797,803,804,810,811,817,818,824,826,831],{"type":47,"value":789},"Resolve the Git-backed scan target, ",{"type":42,"tag":68,"props":791,"children":793},{"className":792},[],[794],{"type":47,"value":795},"repo_name",{"type":47,"value":190},{"type":42,"tag":68,"props":798,"children":800},{"className":799},[],[801],{"type":47,"value":802},"security_scans_dir",{"type":47,"value":190},{"type":42,"tag":68,"props":805,"children":807},{"className":806},[],[808],{"type":47,"value":809},"scan_id",{"type":47,"value":190},{"type":42,"tag":68,"props":812,"children":814},{"className":813},[],[815],{"type":47,"value":816},"scan_dir",{"type":47,"value":160},{"type":42,"tag":68,"props":819,"children":821},{"className":820},[],[822],{"type":47,"value":823},"artifacts_dir",{"type":47,"value":825}," using ",{"type":42,"tag":68,"props":827,"children":829},{"className":828},[],[830],{"type":47,"value":738},{"type":47,"value":75},{"type":42,"tag":178,"props":833,"children":834},{},[835,837,842],{"type":47,"value":836},"Create or adopt the scan goal described in ",{"type":42,"tag":68,"props":838,"children":840},{"className":839},[],[841],{"type":47,"value":606},{"type":47,"value":843}," for that active scan context.",{"type":42,"tag":178,"props":845,"children":846},{},[847,848,854,856,862],{"type":47,"value":472},{"type":42,"tag":68,"props":849,"children":851},{"className":850},[],[852],{"type":47,"value":853},"..\u002F..\u002Freferences\u002Fsecurity-guidance.md",{"type":47,"value":855},", compile the repository's policy to ",{"type":42,"tag":68,"props":857,"children":859},{"className":858},[],[860],{"type":47,"value":861},"\u003Ccontext_dir>\u002Fsecurity_guidance.md",{"type":47,"value":863},", and read it before threat modeling or inspecting source code.",{"type":42,"tag":178,"props":865,"children":866},{},[867,869,874],{"type":47,"value":868},"Run ",{"type":42,"tag":68,"props":870,"children":872},{"className":871},[],[873],{"type":47,"value":530},{"type":47,"value":875}," first.",{"type":42,"tag":665,"props":877,"children":878},{},[879,884],{"type":42,"tag":178,"props":880,"children":881},{},[882],{"type":47,"value":883},"Copy the repository-scoped threat model to the per-scan threat model path without alteration for auditability.",{"type":42,"tag":178,"props":885,"children":886},{},[887],{"type":47,"value":888},"Treat the per-scan threat model path as the source of truth threat model for later phases.",{"type":42,"tag":174,"props":890,"children":892},{"start":891},5,[893],{"type":42,"tag":178,"props":894,"children":895},{},[896,897,902],{"type":47,"value":868},{"type":42,"tag":68,"props":898,"children":900},{"className":899},[],[901],{"type":47,"value":539},{"type":47,"value":903}," as the second step, against the resolved diff and using the per-scan threat model as context.",{"type":42,"tag":665,"props":905,"children":906},{},[907],{"type":42,"tag":178,"props":908,"children":909},{},[910],{"type":47,"value":911},"If discovery produces no technically plausible candidates, stop there, skip validation and attack-path analysis, complete the canonical JSON contract, and finalize the scan.",{"type":42,"tag":174,"props":913,"children":915},{"start":914},6,[916],{"type":42,"tag":178,"props":917,"children":918},{},[919,920,925],{"type":47,"value":868},{"type":42,"tag":68,"props":921,"children":923},{"className":922},[],[924],{"type":47,"value":548},{"type":47,"value":926}," as the third step, for each candidate that came out of discovery.",{"type":42,"tag":665,"props":928,"children":929},{},[930,935],{"type":42,"tag":178,"props":931,"children":932},{},[933],{"type":47,"value":934},"Pass the resolved diff scope, discovery notes, and candidate inventory to validation. Validation should preserve or suppress the provided instances; it should not independently broaden the review into a repository-wide scan.",{"type":42,"tag":178,"props":936,"children":937},{},[938,940,946],{"type":47,"value":939},"Each candidate finding's ",{"type":42,"tag":68,"props":941,"children":943},{"className":942},[],[944],{"type":47,"value":945},"findings\u002F\u003Ccandidate_id>\u002Fcandidate_ledger.jsonl",{"type":47,"value":947}," is part of the validation input. Every candidate finding that came out of discovery must have a discovery receipt before validation starts and a validation receipt before the scan can proceed to final reporting.",{"type":42,"tag":174,"props":949,"children":951},{"start":950},7,[952],{"type":42,"tag":178,"props":953,"children":954},{},[955,956,961],{"type":47,"value":868},{"type":42,"tag":68,"props":957,"children":959},{"className":958},[],[960],{"type":47,"value":557},{"type":47,"value":962}," as the fourth step, for findings that still need reportability, attack-path, and severity analysis after validation.",{"type":42,"tag":665,"props":964,"children":965},{},[966],{"type":42,"tag":178,"props":967,"children":968},{},[969,970,975,977,983],{"type":47,"value":939},{"type":42,"tag":68,"props":971,"children":973},{"className":972},[],[974],{"type":47,"value":945},{"type":47,"value":976}," is part of the attack-path input. Every candidate finding that reaches attack-path analysis must have an attack-path receipt before final reporting, even when the final decision is ",{"type":42,"tag":68,"props":978,"children":980},{"className":979},[],[981],{"type":47,"value":982},"ignore",{"type":47,"value":984},", suppressed, or deferred.",{"type":42,"tag":174,"props":986,"children":988},{"start":987},8,[989],{"type":42,"tag":178,"props":990,"children":991},{},[992,994,1000,1002,1008],{"type":47,"value":993},"Assemble the complete canonical JSON contract last using ",{"type":42,"tag":68,"props":995,"children":997},{"className":996},[],[998],{"type":47,"value":999},"..\u002F..\u002Freferences\u002Ffinal-report.md",{"type":47,"value":1001},"; do not author ",{"type":42,"tag":68,"props":1003,"children":1005},{"className":1004},[],[1006],{"type":47,"value":1007},"report.md",{"type":47,"value":75},{"type":42,"tag":665,"props":1010,"children":1011},{},[1012,1025,1038,1067,1134],{"type":42,"tag":178,"props":1013,"children":1014},{},[1015,1017,1023],{"type":47,"value":1016},"Populate the optional structured details in ",{"type":42,"tag":68,"props":1018,"children":1020},{"className":1019},[],[1021],{"type":47,"value":1022},"..\u002F..\u002Freferences\u002Ffinding-detail-fields.md",{"type":47,"value":1024}," from the same validated evidence used in the generated report.",{"type":42,"tag":178,"props":1026,"children":1027},{},[1028,1030,1036],{"type":47,"value":1029},"For every reportable finding, run ",{"type":42,"tag":68,"props":1031,"children":1033},{"className":1032},[],[1034],{"type":47,"value":1035},"$vulnerability-writeup",{"type":47,"value":1037}," with exactly one dedicated write-up sub-agent. Give it only that finding, its validation and attack-path evidence, relevant source paths and revision, PoC inputs, and the target output directory.",{"type":42,"tag":178,"props":1039,"children":1040},{},[1041,1043,1049,1051,1057,1059,1065],{"type":47,"value":1042},"Write the derived report to ",{"type":42,"tag":68,"props":1044,"children":1046},{"className":1045},[],[1047],{"type":47,"value":1048},"findings\u002F\u003Cslug>\u002F\u003Cslug>.md",{"type":47,"value":1050}," with supporting PoC files under ",{"type":42,"tag":68,"props":1052,"children":1054},{"className":1053},[],[1055],{"type":47,"value":1056},"findings\u002F\u003Cslug>\u002Fpoc\u002F",{"type":47,"value":1058},". Verify the report is a regular file, then set that finding's ",{"type":42,"tag":68,"props":1060,"children":1062},{"className":1061},[],[1063],{"type":47,"value":1064},"writeup.reportPath",{"type":47,"value":1066}," to the matching safe relative path. Do not add the derived report to the sealed artifact list.",{"type":42,"tag":178,"props":1068,"children":1069},{},[1070,1072,1078,1080,1086,1088,1094,1096,1102,1104,1109,1111,1117,1119,1124,1126,1132],{"type":47,"value":1071},"After every write-up is ready, run ",{"type":42,"tag":68,"props":1073,"children":1075},{"className":1074},[],[1076],{"type":47,"value":1077},"$propose-security-hardening",{"type":47,"value":1079}," once over the complete finding collection, detailed write-ups, threat model, coverage, and relevant source. Write its portfolio to ",{"type":42,"tag":68,"props":1081,"children":1083},{"className":1082},[],[1084],{"type":47,"value":1085},"hardening\u002Fhardening.md",{"type":47,"value":1087},", its structured analysis to ",{"type":42,"tag":68,"props":1089,"children":1091},{"className":1090},[],[1092],{"type":47,"value":1093},"hardening\u002Fhardening.json",{"type":47,"value":1095},", and any proposals and diagrams below ",{"type":42,"tag":68,"props":1097,"children":1099},{"className":1098},[],[1100],{"type":47,"value":1101},"hardening\u002F",{"type":47,"value":1103},". Verify ",{"type":42,"tag":68,"props":1105,"children":1107},{"className":1106},[],[1108],{"type":47,"value":1085},{"type":47,"value":1110}," is a regular file, then set ",{"type":42,"tag":68,"props":1112,"children":1114},{"className":1113},[],[1115],{"type":47,"value":1116},"scan.hardening.portfolioPath",{"type":47,"value":1118}," to the fixed relative path ",{"type":42,"tag":68,"props":1120,"children":1122},{"className":1121},[],[1123],{"type":47,"value":1085},{"type":47,"value":1125},". Do not add these derived files to the sealed artifact list. Skip this step and omit ",{"type":42,"tag":68,"props":1127,"children":1129},{"className":1128},[],[1130],{"type":47,"value":1131},"scan.hardening",{"type":47,"value":1133}," when there are no reportable findings.",{"type":42,"tag":178,"props":1135,"children":1136},{},[1137,1139,1144,1146,1152,1154,1160],{"type":47,"value":1138},"Complete the scan once, after all write-ups, hardening guidance, and canonical JSON are ready, so finalization projects the validated JSON and derived-document links into ",{"type":42,"tag":68,"props":1140,"children":1142},{"className":1141},[],[1143],{"type":47,"value":1007},{"type":47,"value":1145},". In the terminal\u002Fchat workflow without ",{"type":42,"tag":68,"props":1147,"children":1149},{"className":1148},[],[1150],{"type":47,"value":1151},"complete_codex_security_scan",{"type":47,"value":1153},", run ",{"type":42,"tag":68,"props":1155,"children":1157},{"className":1156},[],[1158],{"type":47,"value":1159},"python \u003Cplugin_dir>\u002Fscripts\u002Ffinalize_scan_contract.py --scan-dir \u003Cscan_dir> --source-root \u003Crepo_root>",{"type":47,"value":1161}," directly.",{"type":42,"tag":56,"props":1163,"children":1165},{"id":1164},"phase-scope",[1166],{"type":47,"value":1167},"Phase Scope",{"type":42,"tag":665,"props":1169,"children":1170},{},[1171,1183],{"type":42,"tag":178,"props":1172,"children":1173},{},[1174,1176,1182],{"type":47,"value":1175},"Phase 1 (threat model generation) is repository-scope by default, unless the user explicitly asks for narrower scope or provides an authoritative threat model or sufficiently repository-specific security scan guidance such as ",{"type":42,"tag":68,"props":1177,"children":1179},{"className":1178},[],[1180],{"type":47,"value":1181},"AGENTS.md",{"type":47,"value":75},{"type":42,"tag":178,"props":1184,"children":1185},{},[1186],{"type":47,"value":1187},"Phase 2 onward (finding discovery, validation, attack path analysis) are diff-focused and should follow the changed code and its supporting files.",{"type":42,"tag":50,"props":1189,"children":1190},{},[1191],{"type":47,"value":1192},"Treat this asymmetry as intentional:",{"type":42,"tag":665,"props":1194,"children":1195},{},[1196,1201,1206],{"type":42,"tag":178,"props":1197,"children":1198},{},[1199],{"type":47,"value":1200},"use the diff to locate the scan target for later phases",{"type":42,"tag":178,"props":1202,"children":1203},{},[1204],{"type":47,"value":1205},"do not let the diff bias Phase 1 threat model generation, if applicable",{"type":42,"tag":178,"props":1207,"children":1208},{},[1209],{"type":47,"value":1210},"do not let the touched subsystem become the repository threat model unless the user explicitly asks for that narrower scope",{"type":42,"tag":56,"props":1212,"children":1214},{"id":1213},"scan-target",[1215],{"type":47,"value":1216},"Scan Target",{"type":42,"tag":50,"props":1218,"children":1219},{},[1220],{"type":47,"value":1221},"Resolve the exact Git-backed diff before starting:",{"type":42,"tag":665,"props":1223,"children":1224},{},[1225,1235,1240,1245],{"type":42,"tag":178,"props":1226,"children":1227},{},[1228,1230],{"type":47,"value":1229},"PR: compare base branch against current ",{"type":42,"tag":68,"props":1231,"children":1233},{"className":1232},[],[1234],{"type":47,"value":224},{"type":42,"tag":178,"props":1236,"children":1237},{},[1238],{"type":47,"value":1239},"commit: scan the target commit against its parent or requested baseline",{"type":42,"tag":178,"props":1241,"children":1242},{},[1243],{"type":47,"value":1244},"branch diff: scan the requested merge-base to head range",{"type":42,"tag":178,"props":1246,"children":1247},{},[1248],{"type":47,"value":1249},"local patch: scan staged and unstaged working-tree changes against the requested base",{"type":42,"tag":56,"props":1251,"children":1253},{"id":1252},"diff-scoped-discovery",[1254],{"type":47,"value":1255},"Diff-Scoped Discovery",{"type":42,"tag":50,"props":1257,"children":1258},{},[1259,1261,1267],{"type":47,"value":1260},"Use ",{"type":42,"tag":68,"props":1262,"children":1264},{"className":1263},[],[1265],{"type":47,"value":1266},"..\u002Fsecurity-scan\u002Freferences\u002Fscan-artifacts-and-ledger.md",{"type":47,"value":1268}," for the shared scoped file-review, candidate-ledger, subagent, and dedupe rules.",{"type":42,"tag":50,"props":1270,"children":1271},{},[1272],{"type":47,"value":1273},"Diff scans should:",{"type":42,"tag":665,"props":1275,"children":1276},{},[1277,1306,1324,1334,1339],{"type":42,"tag":178,"props":1278,"children":1279},{},[1280,1282,1288,1290,1296,1298,1304],{"type":47,"value":1281},"generate ",{"type":42,"tag":68,"props":1283,"children":1285},{"className":1284},[],[1286],{"type":47,"value":1287},"rank_input.jsonl",{"type":47,"value":1289}," deterministically from changed source-like files with ",{"type":42,"tag":68,"props":1291,"children":1293},{"className":1292},[],[1294],{"type":47,"value":1295},"\u003Cpython_command> \u003Cplugin_dir>\u002Fscripts\u002Fgenerate_rank_input.py make-diff-rank-input --repo \u003Crepo_root> --base \u003Cbase> --mode revisions --head \u003Chead> --out \u003Cdiscovery_dir>\u002Frank_input.jsonl",{"type":47,"value":1297}," for PR, commit, and branch diffs, or ",{"type":42,"tag":68,"props":1299,"children":1301},{"className":1300},[],[1302],{"type":47,"value":1303},"\u003Cpython_command> \u003Cplugin_dir>\u002Fscripts\u002Fgenerate_rank_input.py make-diff-rank-input --repo \u003Crepo_root> --base \u003Cbase> --mode local-patch --out \u003Cdiscovery_dir>\u002Frank_input.jsonl",{"type":47,"value":1305}," for a local patch",{"type":42,"tag":178,"props":1307,"children":1308},{},[1309,1311,1316,1318],{"type":47,"value":1310},"copy every diff row into ",{"type":42,"tag":68,"props":1312,"children":1314},{"className":1313},[],[1315],{"type":47,"value":678},{"type":47,"value":1317}," with ",{"type":42,"tag":68,"props":1319,"children":1321},{"className":1320},[],[1322],{"type":47,"value":1323},"\u003Cpython_command> \u003Cplugin_dir>\u002Fscripts\u002Fgenerate_rank_input.py copy-deep-review-input --rank-input \u003Cdiscovery_dir>\u002Frank_input.jsonl --out \u003Cdiscovery_dir>\u002Fdeep_review_input.jsonl",{"type":42,"tag":178,"props":1325,"children":1326},{},[1327,1329],{"type":47,"value":1328},"deep-review every file in ",{"type":42,"tag":68,"props":1330,"children":1332},{"className":1331},[],[1333],{"type":47,"value":678},{"type":42,"tag":178,"props":1335,"children":1336},{},[1337],{"type":47,"value":1338},"add directly supporting files only when repository evidence shows they are needed to understand the changed security behavior",{"type":42,"tag":178,"props":1340,"children":1341},{},[1342],{"type":47,"value":1343},"stay anchored to the changed code and directly supporting files rather than broadening into unrelated repository-wide enumeration",{"type":42,"tag":56,"props":1345,"children":1347},{"id":1346},"diff-scoped-sibling-coverage",[1348],{"type":47,"value":1349},"Diff-Scoped Sibling Coverage",{"type":42,"tag":50,"props":1351,"children":1352},{},[1353],{"type":47,"value":1354},"For PR, commit, branch, and local-patch scans, stay diff-focused but preserve repeated vulnerable instances that are created or affected by the same changed pattern.",{"type":42,"tag":50,"props":1356,"children":1357},{},[1358],{"type":47,"value":1273},{"type":42,"tag":665,"props":1360,"children":1361},{},[1362,1367,1372,1377,1382,1387,1392],{"type":42,"tag":178,"props":1363,"children":1364},{},[1365],{"type":47,"value":1366},"start from the changed files and the supporting files needed to understand the changed behavior",{"type":42,"tag":178,"props":1368,"children":1369},{},[1370],{"type":47,"value":1371},"expand from a changed route, handler, shared helper, guard, template pattern, query builder, serializer\u002Fdeserializer, filesystem\u002Fnetwork sink, config block, or wrapper to sibling instances that the diff also changes, newly reaches, or affects through the same modified shared dependency",{"type":42,"tag":178,"props":1373,"children":1374},{},[1375],{"type":47,"value":1376},"when the diff adds, removes, or reshapes a guard around an existing parser, deserializer, expression evaluator, filesystem\u002Fpath helper, archive utility, or auth\u002Fauthz helper, use the adjacent pre-existing sink\u002Fcontrol as supporting context for the changed behavior; keep the candidate anchored to the changed guard or newly exposed path unless the user explicitly asks for wider instance expansion",{"type":42,"tag":178,"props":1378,"children":1379},{},[1380],{"type":47,"value":1381},"when a changed wrapper, guard, or API delegates to a shared parser\u002Fdeserializer\u002Fpath\u002Farchive\u002Fauth helper, keep both the wrapper call site and the underlying shared sink\u002Fcontrol line addressable; do not replace the root sink\u002Fcontrol evidence with wrapper-only evidence",{"type":42,"tag":178,"props":1383,"children":1384},{},[1385],{"type":47,"value":1386},"carry each vulnerable sibling instance through discovery and validation with its own affected location, source, closest control, sink, impact, and suppression evidence",{"type":42,"tag":178,"props":1388,"children":1389},{},[1390],{"type":47,"value":1391},"use unchanged siblings as context and negative controls, but report them only when the diff makes them newly vulnerable or changes the shared control or sink they depend on",{"type":42,"tag":178,"props":1393,"children":1394},{},[1395],{"type":47,"value":1396},"stop when the diff-linked pattern family is exhausted, rather than broadening into repository-wide enumeration",{"type":42,"tag":50,"props":1398,"children":1399},{},[1400],{"type":47,"value":1401},"This keeps diff scans precise while avoiding the common failure mode where one representative route or sink hides additional vulnerable siblings introduced by the same patch.",{"type":42,"tag":56,"props":1403,"children":1405},{"id":1404},"final-output",[1406],{"type":47,"value":1407},"Final Output",{"type":42,"tag":50,"props":1409,"children":1410},{},[1411,1413,1418,1420,1426,1428,1434,1436,1441],{"type":47,"value":1412},"Populate all final report semantics in the canonical manifest, findings, and coverage JSON using ",{"type":42,"tag":68,"props":1414,"children":1416},{"className":1415},[],[1417],{"type":47,"value":999},{"type":47,"value":1419},". Generate one detailed ",{"type":42,"tag":68,"props":1421,"children":1423},{"className":1422},[],[1424],{"type":47,"value":1425},"vulnerability-writeup",{"type":47,"value":1427}," for every reportable finding, then run ",{"type":42,"tag":68,"props":1429,"children":1431},{"className":1430},[],[1432],{"type":47,"value":1433},"propose-security-hardening",{"type":47,"value":1435}," once over the complete collection and record the safe derived-document paths. Complete the scan once after both stages; finalization owns ",{"type":42,"tag":68,"props":1437,"children":1439},{"className":1438},[],[1440],{"type":47,"value":1007},{"type":47,"value":1442}," generation. Emit Codex app review directives from the completed canonical findings. Commit scans use this same final-output contract because they are a diff-scan target type.",{"type":42,"tag":56,"props":1444,"children":1446},{"id":1445},"hard-rules",[1447],{"type":47,"value":1448},"Hard Rules",{"type":42,"tag":50,"props":1450,"children":1451},{},[1452,1453,1459],{"type":47,"value":472},{"type":42,"tag":68,"props":1454,"children":1456},{"className":1455},[],[1457],{"type":47,"value":1458},"..\u002F..\u002Freferences\u002Fshared-hard-rules.md",{"type":47,"value":1460}," before applying scan-mode-specific hard rules.",{"type":42,"tag":665,"props":1462,"children":1463},{},[1464,1490],{"type":42,"tag":178,"props":1465,"children":1466},{},[1467,1469,1474,1476,1481,1483,1488],{"type":47,"value":1468},"After any app setup handoff has provided a ",{"type":42,"tag":68,"props":1470,"children":1472},{"className":1471},[],[1473],{"type":47,"value":73},{"type":47,"value":1475},", or in the non-app terminal\u002Fchat workflow, create or adopt the scan goal only after the capability preflight has returned ",{"type":42,"tag":68,"props":1477,"children":1479},{"className":1478},[],[1480],{"type":47,"value":101},{"type":47,"value":1482},", and before substantive scan work. Do not complete it until the resolved diff-scoped files\u002Fworklist rows, candidate ledgers, and final report meet the ",{"type":42,"tag":68,"props":1484,"children":1486},{"className":1485},[],[1487],{"type":47,"value":606},{"type":47,"value":1489}," closure criteria.",{"type":42,"tag":178,"props":1491,"children":1492},{},[1493,1495,1500,1501,1506],{"type":47,"value":1494},"Do not claim diff coverage until every ",{"type":42,"tag":68,"props":1496,"children":1498},{"className":1497},[],[1499],{"type":47,"value":678},{"type":47,"value":680},{"type":42,"tag":68,"props":1502,"children":1504},{"className":1503},[],[1505],{"type":47,"value":686},{"type":47,"value":75},{"items":1508,"total":1713},[1509,1530,1553,1570,1586,1605,1624,1640,1656,1670,1682,1697],{"slug":1510,"name":1510,"fn":1511,"description":1512,"org":1513,"tags":1514,"stars":1527,"repoUrl":1528,"updatedAt":1529},"prior-auth-packet-builder","build healthcare prior authorization packets","Build a concise prior authorization packet from local case files and payer policy docs.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1515,1518,1521,1524],{"name":1516,"slug":1517,"type":15},"Documents","documents",{"name":1519,"slug":1520,"type":15},"Healthcare","healthcare",{"name":1522,"slug":1523,"type":15},"Insurance","insurance",{"name":1525,"slug":1526,"type":15},"Regulatory Compliance","regulatory-compliance",28169,"https:\u002F\u002Fgithub.com\u002Fopenai\u002Fopenai-agents-python","2026-04-16T05:11:39.180399",{"slug":1531,"name":1531,"fn":1532,"description":1533,"org":1534,"tags":1535,"stars":1550,"repoUrl":1551,"updatedAt":1552},"aspnet-core","build ASP.NET Core web applications","Build, review, refactor, or architect ASP.NET Core web applications using current official guidance for .NET web development. Use when working on Blazor Web Apps, Razor Pages, MVC, Minimal APIs, controller-based Web APIs, SignalR, gRPC, middleware, dependency injection, configuration, authentication, authorization, testing, performance, deployment, or ASP.NET Core upgrades.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1536,1539,1541,1544,1547],{"name":1537,"slug":1538,"type":15},".NET","dotnet",{"name":1540,"slug":1531,"type":15},"ASP.NET Core",{"name":1542,"slug":1543,"type":15},"Blazor","blazor",{"name":1545,"slug":1546,"type":15},"C#","csharp",{"name":1548,"slug":1549,"type":15},"Web Development","web-development",23787,"https:\u002F\u002Fgithub.com\u002Fopenai\u002Fskills","2026-04-12T05:07:02.819491",{"slug":1554,"name":1554,"fn":1555,"description":1556,"org":1557,"tags":1558,"stars":1550,"repoUrl":1551,"updatedAt":1569},"chatgpt-apps","build ChatGPT Apps SDK applications","Build, scaffold, refactor, and troubleshoot ChatGPT Apps SDK applications that combine an MCP server and widget UI. Use when Codex needs to design tools, register UI resources, wire the MCP Apps bridge or ChatGPT compatibility APIs, apply Apps SDK metadata or CSP or domain settings, or produce a docs-aligned project scaffold. Prefer a docs-first workflow by invoking the openai-docs skill or OpenAI developer docs MCP tools before generating code.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1559,1562,1565,1568],{"name":1560,"slug":1561,"type":15},"Apps SDK","apps-sdk",{"name":1563,"slug":1564,"type":15},"ChatGPT","chatgpt",{"name":1566,"slug":1567,"type":15},"MCP","mcp",{"name":9,"slug":8,"type":15},"2026-04-12T05:07:05.468097",{"slug":1571,"name":1571,"fn":1572,"description":1573,"org":1574,"tags":1575,"stars":1550,"repoUrl":1551,"updatedAt":1585},"cli-creator","build CLIs from API docs","Build a composable CLI for Codex from API docs, an OpenAPI spec, existing curl examples, an SDK, a web app, an admin tool, or a local script. Use when the user wants Codex to create a command-line tool that can run from any repo, expose composable read\u002Fwrite commands, return stable JSON, manage auth, and pair with a companion skill.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1576,1579,1582],{"name":1577,"slug":1578,"type":15},"API Development","api-development",{"name":1580,"slug":1581,"type":15},"CLI","cli",{"name":1583,"slug":1584,"type":15},"Codex","codex","2026-04-12T05:07:04.132762",{"slug":1587,"name":1587,"fn":1588,"description":1589,"org":1590,"tags":1591,"stars":1550,"repoUrl":1551,"updatedAt":1604},"cloudflare-deploy","deploy projects to Cloudflare","Deploy applications and infrastructure to Cloudflare using Workers, Pages, and related platform services. Use when the user asks to deploy, host, publish, or set up a project on Cloudflare.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1592,1595,1598,1601],{"name":1593,"slug":1594,"type":15},"Cloudflare","cloudflare",{"name":1596,"slug":1597,"type":15},"Cloudflare Pages","cloudflare-pages",{"name":1599,"slug":1600,"type":15},"Cloudflare Workers","cloudflare-workers",{"name":1602,"slug":1603,"type":15},"Deployment","deployment","2026-04-12T05:07:14.275118",{"slug":1606,"name":1606,"fn":1607,"description":1608,"org":1609,"tags":1610,"stars":1550,"repoUrl":1551,"updatedAt":1623},"define-goal","define and set measurable project goals","Help the user define a concrete, measurable goal before starting work, especially when they ask to use the goal tool, create a goal, set an objective, clarify success criteria, or turn a fuzzy intention into a quantitative outcome. Use this skill for goal creation and goal refinement only; it does not manage durable snapshots, decision logs, or long-running execution artifacts.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1611,1614,1617,1620],{"name":1612,"slug":1613,"type":15},"Productivity","productivity",{"name":1615,"slug":1616,"type":15},"Project Management","project-management",{"name":1618,"slug":1619,"type":15},"Strategy","strategy",{"name":1621,"slug":1622,"type":15},"Task Management","task-management","2026-05-23T06:17:16.870838",{"slug":1625,"name":1625,"fn":1626,"description":1627,"org":1628,"tags":1629,"stars":1550,"repoUrl":1551,"updatedAt":1639},"figma","translate Figma designs into code","Use the Figma MCP server to fetch design context, screenshots, variables, and assets from Figma, and to translate Figma nodes into production code. Trigger when a task involves Figma URLs, node IDs, design-to-code implementation, or Figma MCP setup and troubleshooting.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1630,1633,1635,1638],{"name":1631,"slug":1632,"type":15},"Design","design",{"name":1634,"slug":1625,"type":15},"Figma",{"name":1636,"slug":1637,"type":15},"Frontend","frontend",{"name":1566,"slug":1567,"type":15},"2026-04-12T05:06:47.939943",{"slug":1641,"name":1641,"fn":1642,"description":1643,"org":1644,"tags":1645,"stars":1550,"repoUrl":1551,"updatedAt":1655},"figma-code-connect-components","connect Figma designs to code components","Connects Figma design components to code components using Code Connect mapping tools. Use when user says \"code connect\", \"connect this component to code\", \"map this component\", \"link component to code\", \"create code connect mapping\", or wants to establish mappings between Figma designs and code implementations. For canvas writes via `use_figma`, use `figma-use`.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1646,1647,1650,1651,1652],{"name":1631,"slug":1632,"type":15},{"name":1648,"slug":1649,"type":15},"Design System","design-system",{"name":1634,"slug":1625,"type":15},{"name":1636,"slug":1637,"type":15},{"name":1653,"slug":1654,"type":15},"UI Components","ui-components","2026-05-10T05:59:52.971881",{"slug":1657,"name":1657,"fn":1658,"description":1659,"org":1660,"tags":1661,"stars":1550,"repoUrl":1551,"updatedAt":1669},"figma-create-design-system-rules","generate design system rules from Figma","Generates custom design system rules for the user's codebase. Use when user says \"create design system rules\", \"generate rules for my project\", \"set up design rules\", \"customize design system guidelines\", or wants to establish project-specific conventions for Figma-to-code workflows. Requires Figma MCP server connection.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1662,1663,1664,1667,1668],{"name":1631,"slug":1632,"type":15},{"name":1648,"slug":1649,"type":15},{"name":1665,"slug":1666,"type":15},"Documentation","documentation",{"name":1634,"slug":1625,"type":15},{"name":1636,"slug":1637,"type":15},"2026-05-16T06:07:47.821474",{"slug":1671,"name":1671,"fn":1672,"description":1673,"org":1674,"tags":1675,"stars":1550,"repoUrl":1551,"updatedAt":1681},"figma-implement-design","translate Figma designs into application code","Translates Figma designs into production-ready application code with 1:1 visual fidelity. Use when implementing UI code from Figma files, when user mentions \"implement design\", \"generate code\", \"implement component\", provides Figma URLs, or asks to build components matching Figma specs. For Figma canvas writes via `use_figma`, use `figma-use`.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1676,1677,1678,1679,1680],{"name":1631,"slug":1632,"type":15},{"name":1634,"slug":1625,"type":15},{"name":1636,"slug":1637,"type":15},{"name":1653,"slug":1654,"type":15},{"name":1548,"slug":1549,"type":15},"2026-05-16T06:07:40.583615",{"slug":1683,"name":1683,"fn":1684,"description":1685,"org":1686,"tags":1687,"stars":1550,"repoUrl":1551,"updatedAt":1696},"hatch-pet","create animated pets for Codex","Create, repair, validate, visually QA, and package Codex-compatible animated pets and pet spritesheets from character art, generated images, company or prospect brand cues, or visual references. Use when a user wants a lightweight-worker Codex pet workflow, a non-pixel custom pet style, a prospect or company mascot pet, or a full 8x9 animated pet atlas with transparent unused cells, QA contact sheets, and pet.json packaging. This skill composes the installed $imagegen system skill for visual generation and uses bundled scripts for deterministic spritesheet assembly.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1688,1691,1692,1695],{"name":1689,"slug":1690,"type":15},"Animation","animation",{"name":1583,"slug":1584,"type":15},{"name":1693,"slug":1694,"type":15},"Creative","creative",{"name":1631,"slug":1632,"type":15},"2026-05-02T05:31:48.48485",{"slug":1698,"name":1698,"fn":1699,"description":1700,"org":1701,"tags":1702,"stars":1550,"repoUrl":1551,"updatedAt":1712},"imagegen","generate and edit raster images","Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output should be a bitmap asset rather than repo-native code or vector. Do not use when the task is better handled by editing existing SVG\u002Fvector\u002Fcode-native assets, extending an established icon or logo system, or building the visual directly in HTML\u002FCSS\u002Fcanvas.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1703,1704,1705,1708,1711],{"name":1693,"slug":1694,"type":15},{"name":1631,"slug":1632,"type":15},{"name":1706,"slug":1707,"type":15},"Image Generation","image-generation",{"name":1709,"slug":1710,"type":15},"Images","images",{"name":9,"slug":8,"type":15},"2026-05-15T06:23:24.312127",675,{"items":1715,"total":1828},[1716,1732,1748,1760,1778,1796,1816],{"slug":1717,"name":1717,"fn":1718,"description":1719,"org":1720,"tags":1721,"stars":25,"repoUrl":26,"updatedAt":27},"accessibility-and-inclusive-visualization","make data visualizations accessible","Make data visualizations accessible and inclusive. Use when the user needs chart or diagram accessibility guidance, text alternatives for complex visuals, color and contrast review, keyboard support, reduced-motion behavior for animation or parallax, or an accessibility QA workflow for exported figures, UML-like diagrams, and dashboards.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1722,1725,1728,1731],{"name":1723,"slug":1724,"type":15},"Accessibility","accessibility",{"name":1726,"slug":1727,"type":15},"Charts","charts",{"name":1729,"slug":1730,"type":15},"Data Visualization","data-visualization",{"name":1631,"slug":1632,"type":15},{"slug":1733,"name":1733,"fn":1734,"description":1735,"org":1736,"tags":1737,"stars":25,"repoUrl":26,"updatedAt":1747},"agent-browser","automate browser interactions for agents","Browser automation CLI for AI agents. Use when the user needs to interact with websites, verify dev server output, test web apps, navigate pages, fill forms, click buttons, take screenshots, extract data, or automate any browser task. Also triggers when a dev server starts so you can verify it visually.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1738,1741,1744],{"name":1739,"slug":1740,"type":15},"Agents","agents",{"name":1742,"slug":1743,"type":15},"Browser Automation","browser-automation",{"name":1745,"slug":1746,"type":15},"Testing","testing","2026-04-06T18:41:03.44016",{"slug":1749,"name":1749,"fn":1750,"description":1751,"org":1752,"tags":1753,"stars":25,"repoUrl":26,"updatedAt":1759},"agent-browser-verify","verify dev server output with automated browser","Automated browser verification for dev servers. Triggers when a dev server starts to run a visual gut-check with agent-browser — verifies the page loads, checks for console errors, validates key UI elements, and reports pass\u002Ffail before continuing.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1754,1755,1758],{"name":1742,"slug":1743,"type":15},{"name":1756,"slug":1757,"type":15},"Local Development","local-development",{"name":1745,"slug":1746,"type":15},"2026-04-06T18:41:17.526867",{"slug":1761,"name":1761,"fn":1762,"description":1763,"org":1764,"tags":1765,"stars":25,"repoUrl":26,"updatedAt":1777},"agents-sdk","build AI agents on Cloudflare Workers","Build AI agents on Cloudflare Workers using the Agents SDK. Load when creating stateful agents, durable workflows, real-time WebSocket apps, scheduled tasks, MCP servers, or chat applications. Covers Agent class, state management, callable RPC, Workflows integration, and React hooks. Biases towards retrieval from Cloudflare docs over pre-trained knowledge.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1766,1767,1768,1771,1774],{"name":1739,"slug":1740,"type":15},{"name":1599,"slug":1600,"type":15},{"name":1769,"slug":1770,"type":15},"SDK","sdk",{"name":1772,"slug":1773,"type":15},"Serverless","serverless",{"name":1775,"slug":1776,"type":15},"WebSockets","websockets","2026-04-06T18:39:51.717063",{"slug":1779,"name":1779,"fn":1780,"description":1781,"org":1782,"tags":1783,"stars":25,"repoUrl":26,"updatedAt":1795},"ai-elements","build chat UIs with AI Elements","AI Elements component library guidance — pre-built React components for AI interfaces built on shadcn\u002Fui. Use when building chat UIs, message displays, tool call rendering, streaming responses, reasoning panels, or any AI-native interface with the AI SDK.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1784,1785,1788,1791,1792],{"name":1636,"slug":1637,"type":15},{"name":1786,"slug":1787,"type":15},"React","react",{"name":1789,"slug":1790,"type":15},"shadcn\u002Fui","shadcn-ui",{"name":1653,"slug":1654,"type":15},{"name":1793,"slug":1794,"type":15},"Vercel","vercel","2026-04-06T18:40:59.619419",{"slug":1797,"name":1797,"fn":1798,"description":1799,"org":1800,"tags":1801,"stars":25,"repoUrl":26,"updatedAt":1815},"ai-gateway","configure Vercel AI Gateway","Vercel AI Gateway expert guidance. Use when configuring model routing, provider failover, cost tracking, or managing multiple AI providers through a unified API.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1802,1805,1808,1811,1814],{"name":1803,"slug":1804,"type":15},"AI Infrastructure","ai-infrastructure",{"name":1806,"slug":1807,"type":15},"Cost Optimization","cost-optimization",{"name":1809,"slug":1810,"type":15},"LLM","llm",{"name":1812,"slug":1813,"type":15},"Performance","performance",{"name":1793,"slug":1794,"type":15},"2026-04-06T18:40:44.377464",{"slug":1817,"name":1817,"fn":1818,"description":1819,"org":1820,"tags":1821,"stars":25,"repoUrl":26,"updatedAt":1827},"ai-generation-persistence","implement persistence patterns for AI generations","AI generation persistence patterns — unique IDs, addressable URLs, database storage, and cost tracking for every LLM generation",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[1822,1823,1826],{"name":1806,"slug":1807,"type":15},{"name":1824,"slug":1825,"type":15},"Database","database",{"name":1809,"slug":1810,"type":15},"2026-04-06T18:41:08.513425",600]