project-init

Codex compatibility note:

• Invoke repository skills with $skill-name in Codex; this mirrored copy rewrites legacy Claude /skill-name references.
• Task tracker mandate: BEFORE executing any workflow or skill step, create/update task tracking for all steps and keep it synchronized as progress changes.
• User-question prompts mean to ask the user directly in Codex.
• Ignore Claude-specific mode-switch instructions when they appear.
• Strict execution contract: when a user explicitly invokes a skill, execute that skill protocol as written.
• Subagent authorization: when a skill is user-invoked or AI-detected and its protocol requires subagents, that skill activation authorizes use of the required spawn_agent subagent(s) for that task.
• Do not skip, reorder, or merge protocol steps unless the user explicitly approves the deviation first.
• For workflow skills, execute each listed child-skill step explicitly and report step-by-step evidence.
• If a required step/tool cannot run in this environment, stop and ask the user before adapting.

Codex Project-Reference Loading (No Hooks)

Codex uses static project-reference loading instead of runtime-injected project docs. When coding, planning, debugging, testing, or reviewing, open project docs explicitly using this routing.

Always read:

• docs/project-config.json (project-specific paths, commands, modules, and workflow/test settings)
• docs/project-reference/docs-index-reference.md (routes to the full docs/project-reference/* catalog)
• docs/project-reference/lessons.md (always-on guardrails and anti-patterns)

Missing/stale context route: If docs/project-config.json, the docs index, lessons.md, CLAUDE.md, AGENTS.md, or any task-required reference doc is missing or stale, auto-run $project-init or the narrow setup route ($project-config, $docs-init, $scan-all, $scan --target=<key>, $claude-md-init) before ordinary project-specific work. If Codex mirrors or AGENTS.md are missing/stale, ask the user to run $sync-codex; do not auto-run it.

Situation-based docs:

• Project structure/architecture/tech-stack/deployment/setup (any layer — backend, frontend, or infra): project-structure-reference.md
• Backend/CQRS/API/domain/entity changes: backend-patterns-reference.md, domain-entities-reference.md
• Frontend/UI/styling/design-system: frontend-patterns-reference.md, scss-styling-guide.md, design-system/README.md
• Spec authoring, docs/specs/ pathing, or TC format: feature-spec-reference.md, spec-system-reference.md, spec-principles.md
• Behavior/public-contract changes or spec-test-code sync: workflow-spec-test-code-cycle-reference.md plus the spec docs above
• Derived spec indexes/ERDs/reimplementation guides: spec-system-reference.md and source Feature Specs under docs/specs/
• Integration test implementation/review: integration-test-reference.md
• E2E test implementation/review: e2e-test-reference.md
• Code review/audit work: code-review-rules.md plus domain docs above based on changed files

Do not read all docs blindly. Start from docs-index-reference.md, then open only relevant files for the task.

[IMPORTANT] Create complete task plan before shell checks, scans, generators, edits, or skill calls. [IMPORTANT] After config exists, run $scan-all + $workflow-code-to-spec as post-config barrier; do not proceed until both return outcome/blocker/deferral. [IMPORTANT] After setup/review/verification is otherwise done, spawn background $graph-build sub-agent and record outcome.

Quick Summary

Goal: Initialize or re-evaluate all portable project context files through one idempotent route, so any agent — with or without hooks — can discover missing context and reach a verified state with project config, reference docs, root instruction files, and Codex mirrors aligned.

Workflow:

1. Plan Tasks - Before any setup action, create task-tracking entries for every project-init phase and every required final skill call.
2. Assess - Classify folder state and current context-file health.
3. Bootstrap - Create missing config/doc stubs when useful, then route scans/generators.
4. Populate - After project config is initialized/refreshed, start the post-config parallel group: call $scan-all and $workflow-code-to-spec.
5. Spec Finalization - Resolve the mandatory docs/specs gate from $workflow-code-to-spec; only empty/no-content or no-accepted-capability projects may defer with evidence.
6. Review - Run $changes-review, then $why-review as final quality gates after scan/spec completion.
7. Verify - Validate config, root files, docs, mirrors, spec workflow status, and staleness.
8. Background Graph Refresh - After setup/review/verification is otherwise done, spawn a background sub-agent to run $graph-build.
9. Report - List completed actions, skipped actions, and remaining manual steps.

Key Rules:

• MUST ATTENTION run this before ordinary work when required config/docs/root instruction files are missing.
• MUST ATTENTION preserve user-authored CLAUDE.md and AGENTS.md; use smart-merge/update paths, never blind overwrite.
• MUST ATTENTION keep reusable skill text project-neutral; local rules belong in project config/reference docs.
• MUST ATTENTION use configured portability paths from .claude/.ck.json when present.
• MUST ATTENTION before any shell check, scan, generator, or file edit, create a complete task plan covering assessment, setup routes, final skill calls, verification, report, and lessons.
• MUST ATTENTION immediately after $project-config initializes or refreshes config, create and execute a post-config parallel task group with Call $scan-all and Call $workflow-code-to-spec.
• MUST ATTENTION call $scan-all after config initialization for every content-bearing project; skip only empty/no-content projects with recorded evidence.
• MUST ATTENTION invoke $workflow-code-to-spec for every content-bearing project; it may run in parallel with $scan-all after config exists, but do not treat "handoff suggested" as completion.
• MUST ATTENTION end every setup run with final task-plan rows, in order: Call $changes-review, Call $why-review, Spawn background $graph-build sub-agent, after the scan/spec parallel group is resolved.
• MUST ATTENTION run $graph-build as a required final background sub-agent task after setup/review/verification work is otherwise done; do not run this final graph refresh inline in the main context.
• MUST ATTENTION ask the user to run $sync-codex or its standalone node runner when Codex mirrors/root files are missing or stale; do not auto-run this user-invoked-only sync route.

Scope

$project-init is the canonical coordinator for portable setup. It does not replace lower-level skills; it decides which one to run and when:

Project-init test matrix — Defines setup states, routing expectations, and regression cases for post-config scan/spec and final graph behavior. MUST ATTENTION read references/use-cases-and-test-cases.md when creating plans, tests, or reviewing changes to this setup.

Phase -1: Required Task Plan

Before Phase 0 shell checks, create a full task-tracking plan. The plan MUST include many small, observable tasks and MUST NOT start execution until these rows exist.

Minimum required task rows:

1. Read project-init instructions and setup reference files.
2. Assess folder classification, config, docs, root files, mirrors, graph, and spec inventory.
3. Run required config setup route ($project-config) or mark skipped with evidence.
4. Create a post-config parallel group with both Call $scan-all and Call $workflow-code-to-spec as sibling tasks.
5. Call $scan-all after config initialization or mark skipped only for empty/no-content evidence.
6. Call $workflow-code-to-spec after config initialization and complete its Step 0 mode/scope route, or record the exact blocking user confirmation needed.
7. Wait at a barrier until both post-config parallel tasks are completed, blocked, or evidence-deferred.
8. Run required root-instruction route ($claude-md-init) or mark skipped with evidence.
9. Resolve Codex mirror route by asking for $sync-codex when required.
10. Call $changes-review after the scan/spec barrier.
11. Call $why-review after $changes-review.
12. Run verification commands.
13. Spawn a background sub-agent task named Spawn background $graph-build sub-agent after setup/review/verification is otherwise done.
14. Record the background $graph-build sub-agent outcome or explicit blocker.
15. Report files changed, routes invoked, scan/spec outcomes, review outcomes, background graph outcome, verification output, and remaining manual actions.
16. Analyze AI mistakes and reusable lessons.

Keep exactly one row in_progress. Mark each row completed immediately after its evidence is recorded.

Phase 0: Assess State

Use shell checks, not memory. Record evidence for every state claim:

node -e "const h=require('./.claude/hooks/lib/session-init-helpers.cjs'); console.log(JSON.stringify({hasProjectContent:h.hasProjectContent(), isGreenfield:h.isGreenfieldProject()}, null, 2))"
node -e "const s=require('./.claude/hooks/lib/session-init-helpers.cjs'); console.log(JSON.stringify(s.checkProjectConfig(), null, 2))"
node -e "const a=require('./.claude/hooks/lib/agent-files-state.cjs'); console.log(JSON.stringify(a.getAgentFileIssues(), null, 2))"

Also check:

• Config path: node -e "console.log(require('./.claude/hooks/lib/project-config-loader.cjs').getConfiguredProjectConfigPath())"
• Docs index path: node -e "console.log(require('./.claude/hooks/lib/project-config-loader.cjs').getConfiguredDocsIndexPath())"
• Feature docs path: node -e "console.log('docs/specs/')"
• Placeholder docs: use isPlaceholderFile() from .claude/hooks/lib/session-init-helpers.cjs.
• Stale docs: use getStaleReferenceDocs() from .claude/hooks/lib/session-init-helpers.cjs.
• Reference-doc canonical-floor drift: node -e "const h=require('./.claude/hooks/lib/session-init-helpers.cjs');const{loadProjectConfig}=require('./.claude/hooks/lib/project-config-loader.cjs');const r=h.normalizeReferenceDocs((loadProjectConfig()||{}).referenceDocs);console.log(JSON.stringify({changed:r.changed,renames:r.renames,added:r.added,removedLegacy:r.removedLegacy},null,2))". changed:true (non-empty renames/added/removedLegacy) means the config drifted below the canonical reference-doc floor (legacy filenames, missing canonical entries, or wrong order) and MUST be repaired in Phase 2 step 1a before the scan/spec barrier.
• Spec inventory: probe the fixed feature docs path: node -e "const p='docs/specs/'; const fs=require('fs'); console.log(JSON.stringify({path:p, exists:fs.existsSync(p)}, null, 2))"

Phase 1: Decide Route

Phase 2: Execute Order

Run phases sequentially except the explicit post-config parallel group. After each phase, re-run Phase 0 checks.

1. Config - $project-config when config is missing, skeleton, invalid, or stale relative to the workspace. 1a. Reference-doc normalization — canonical floor (MANDATORY when Phase 0 probe reports changed:true) - repair reference-doc drift BEFORE the scan/spec barrier so every project converges to the framework floor regardless of starting state (no docs / partial / legacy names / wrong standard):
   • Rewrite config.referenceDocs to normalizeReferenceDocs(config.referenceDocs).normalized (canonical order, legacy names resolved via the alias map, canonical templatePaths preserved, genuine project-specific extras kept). Never delete or rename a canonical entry.
   • For each renames[] {from,to}: git mv docs/project-reference/<from> docs/project-reference/<to> when <to> is absent; if <to> already exists, <from> is a stale duplicate → confirm <to> holds the canonical content, then git rm docs/project-reference/<from>. Migrate every downstream textual reference (docs-index-reference.md, project-structure-reference.md) <from> → <to>, then ask the user to re-run $sync-codex so the regenerated mirrors match.
   • added[] canonical docs missing on disk are created idempotently by the SessionStart session-init-docs.cjs hook (or $scan --target=<key>) from DEFAULT_REFERENCE_DOCS + templatePath; do not hand-fabricate their content.
   • Re-run the Phase 0 normalize probe and proceed only when it reports changed:false with empty renames/added/removedLegacy.
2. Post-config parallel context build (MANDATORY) - after config exists, create sibling tasks Call $scan-all and Call $workflow-code-to-spec; run them in parallel when the environment/tooling supports parallel skill work, otherwise execute both before crossing the barrier.
   • $scan-all: required for content-bearing projects after config initialization. Skip only empty/no-content projects with evidence. Use $docs-init or targeted $scan --target=<key> only as follow-up repair when scan output proves it is needed.
   • $workflow-code-to-spec: required for content-bearing projects after config initialization. It may complete Step 0/mode selection in parallel with $scan-all; if it needs scan results for capability enumeration, pause inside that workflow until $scan-all evidence is available.
   • Barrier: do not proceed to root instruction updates, mirrors, final review, verification, or report until both sibling tasks have an outcome: completed, explicit blocker, or evidence-backed deferral.
3. Root instructions - $claude-md-init --mode init|update.
4. Codex mirror - ask the user to run $sync-codex or node .claude/skills/sync-codex/scripts/run-codex-sync.mjs.
5. Spec workflow outcome (MANDATORY) - confirm the existing Call $workflow-code-to-spec task has one of these outcomes before review:
   • Empty/no-content folder: do not deep-scan or fabricate capabilities; mark this task Deferred: no project content or accepted capability scope, and report the next trigger ($workflow-idea-to-spec or $greenfield, then $workflow-code-to-spec init-full).
   • Greenfield with accepted product/capability scope or real code scaffold: invoke $workflow-code-to-spec, suggest init-full, and let its Step 0 confirm mode/bucket/capability.
   • Existing/grown project with docs/specs/ missing or empty: invoke $workflow-code-to-spec, suggest init-full, and require divide-and-conquer grouping when capability count is large.
   • Existing/grown project with Feature Specs already present: invoke $workflow-code-to-spec, suggest audit; suggest update when git diff or a new requirement/PBI is the trigger.
6. Review changes (MANDATORY) - create a final task named Call $changes-review, invoke $changes-review, and record pass/fail or explicit blocker.
7. Why review (MANDATORY) - create a final task named Call $why-review, invoke $why-review after $changes-review, and record pass/fail or explicit blocker.
8. Enhance - $prompt-enhance on newly created or heavily updated skill/docs files.
9. Queue background graph refresh (MANDATORY FINAL) - create a final task named Spawn background $graph-build sub-agent; do not execute it until Phase 4 verification and setup/review work are otherwise done.

Phase 2.5: Mandatory Spec Workflow Finalization

This phase starts as the Call $workflow-code-to-spec sibling task in the post-config parallel group. It is ALWAYS represented in task tracking and MUST be resolved before $changes-review, $why-review, final verification, and report.

Do not mark $project-init complete while this phase is unresolved. A valid resolution is either: (1) $workflow-code-to-spec invoked and its Step 0/mode route completed, (2) $workflow-code-to-spec invoked and blocked on explicit user confirmation of mode/bucket/capability, or (3) explicit deferral with evidence that the project has no content/capability source yet. A plain recommendation or handoff without invoking $workflow-code-to-spec is NOT a valid resolution.

$scan-all and $workflow-code-to-spec are allowed to run in parallel only after project config exists. The project-init report must include both outcomes and must state whether $workflow-code-to-spec consumed scan evidence directly or paused pending scan output.

Phase 2.6: Mandatory Final Review Skills

After Phase 2.5, always create and execute these final tasks in order:

1. Call $changes-review - run after $workflow-code-to-spec so the setup/spec changes are reviewed from the current diff.
2. Call $why-review - run after $changes-review to validate rationale and avoid closing on unchallenged setup decisions.

If either skill cannot run because the environment lacks the required tool, stop and report the missing tool. Do not silently replace them with a summary.

Phase 3: Hookless Agent Rule

For Codex or any environment without Claude hooks:

• If docs/project-config.json, the docs index, lessons.md, CLAUDE.md, AGENTS.md, or a task-required reference doc is missing or stale, invoke $project-init before ordinary task work.
• If $project-init cannot run because required tools are absent, report the missing tool and the exact lower-level route that remains.
• Do not proceed with project-specific coding from guessed context.

Phase 4: Verification

Minimum verification before declaring setup complete:

node -e "const{validateConfig,formatResult}=require('./.claude/hooks/lib/project-config-schema.cjs');const{getConfiguredProjectConfigPath}=require('./.claude/hooks/lib/project-config-loader.cjs');const c=JSON.parse(require('fs').readFileSync(getConfiguredProjectConfigPath(),'utf-8'));console.log(formatResult(validateConfig(c)))"
node -e "const a=require('./.claude/hooks/lib/agent-files-state.cjs'); console.log(JSON.stringify(a.getAgentFileIssues(), null, 2))"
node .claude/hooks/tests/test-all-hooks.cjs
node .claude/skills/skill-creator/scripts/validate-skills.cjs --path .claude/skills/project-init

Spec workflow verification before declaring setup complete:

• Confirm the task list contains a post-config parallel group with Call $scan-all and Call $workflow-code-to-spec.
• Confirm Call $scan-all ran after $project-config, or was skipped only with empty/no-content evidence.
• Confirm the task list contains Call $changes-review and Call $why-review as the final review skill-call rows after the scan/spec barrier.
• Confirm the $workflow-code-to-spec outcome is one of: init-full, audit, update, blocked on user-confirmed mode/bucket/capability, or an evidence-backed deferral for empty/no-capability projects.
• Confirm $changes-review ran after $workflow-code-to-spec, or stopped with an explicit missing-tool blocker.
• Confirm $why-review ran after $changes-review, or stopped with an explicit missing-tool blocker.
• Confirm Spawn background $graph-build sub-agent ran after setup/review/verification was otherwise done, or stopped with an explicit missing-tool/dependency blocker.
• Confirm the Feature Spec root is the fixed path docs/specs/.
• Confirm the hook-independent Workflow-First Gate (<!-- CK:WORKFLOW-GATE --> block) is present at the TOP of CLAUDE.md and AGENTS.md so routing survives without hooks. If missing, re-run $claude-md-init (CLAUDE.md), then ask the user to run $sync-codex (AGENTS.md mirror).
• For grown projects, confirm large scope is split per $workflow-code-to-spec (>10 capabilities grouped; 4-10 capabilities sub-agented).

If Codex mirrors were changed, run:

node .claude/skills/sync-codex/scripts/run-codex-sync.mjs --only=tests,wf-cycle,sk-proto,residue,sdd

Phase 5: Mandatory Final Background Graph Build

After Phase 4 verification and setup/review work are otherwise done, always execute this final task:

1. Spawn background $graph-build sub-agent - spawn a background sub-agent whose only job is to invoke $graph-build with default auto-detect scope, read its result, and return a concise outcome.

Rules:

• Run this final graph refresh in a sub-agent/background task, not inline in the main context.
• The task is required even if an earlier graph check found .code-graph/graph.db; existing graph presence changes $graph-build from full build to auto-detected update, not a skip.
• Keep the task open until the background sub-agent returns, or record an explicit blocker such as missing Python/dependencies.
• The project-init report must include the background graph task outcome.

Output

Report:

• Folder classification: empty, greenfield, existing, or already initialized.
• Files created/updated/skipped: project config, reference docs, CLAUDE.md, AGENTS.md, mirrors.
• Lower-level skills/scripts invoked.
• Post-config parallel skill calls: $scan-all and $workflow-code-to-spec, each with outcome and evidence.
• Final review skill calls: $changes-review, $why-review, each with outcome and evidence.
• Final background graph call: $graph-build sub-agent outcome, scope/build type, and blocker if any.
• Spec workflow finalization: invoked mode (init-full, audit, update), user-confirmation blocker, or exact deferral reason and next trigger.
• Verification commands and results.
• Remaining manual action, especially any user-confirmed $sync-codex step.

Parallel Sub-Agent Dispatch — Plan parallelism the moment a task breakdown exists, BEFORE executing it — running provably independent tasks sequentially wastes wall-clock. Applies to every multi-step job: workflow steps, planning, batch updates, investigation, research, scans, reviews, doc sync. Plan execution is metadata-gated, NEVER default-parallel — fan-out follows ONLY what the plan declares (PAR/SEQ tags + per-phase write set); an untagged plan runs sequentially — why: a derived write set cannot see cascade or generated writes.

1. Tag every task PAR or SEQ. PAR = inputs exclude every pending task's output AND write set disjoint from every other PAR. Else SEQ — MUST ATTENTION name the dependency forcing it.
2. Group PAR into waves. No edge between members. Two writers of one file NEVER share a wave. Read-only work (search, investigation, review, research) parallelizes freely.
3. Declare before dispatch: Parallel plan: wave 1 = [...] · wave 2 = [...] · SEQ = [...] (reason).
4. Spawn each wave in ONE message — every spawn_agent call in one response, NEVER dripped per turn. Route each task to its specialist (.claude/skills/shared/sub-agent-selection-guide.md); NEVER code-reviewer as catch-all.
5. Brief each sub-agent self-contained: goal · scope + owned files · reference docs · return contract (summary + Full report: path, per SYNC:subagent-return-contract) · incremental persistence to plans/reports/ (per SYNC:incremental-persistence).
6. Barrier per wave. Advance ONLY after EVERY member returns (a skipped conditional counts as returned). Merge, mark each task completed/skipped, THEN dispatch the next wave. Mutating steps wait for the barrier.
7. One level deep. A dispatched sub-agent executes its own brief; further fan-out stays the orchestrator's job unless that agent's .claude/agents/*.md definition authorizes it.

NEVER parallelize: tasks sharing a write target · a task consuming a pending task's output · trivial single-file work (dispatch overhead > gain) · an order a skill or workflow explicitly fixes · gates awaiting user approval.

Blocked until: MUST ATTENTION every task tagged PAR/SEQ with a named reason per SEQ · waves declared + write-set disjointness checked · each wave spawned in ONE message · barrier honored before the next wave.

• MANDATORY After planning tasks, tag each PAR/SEQ and spawn every PAR wave as parallel sub-agents in ONE message — default parallel for workflows, batch updates, investigation, research, reviews; plan execution fans out ONLY on what the plan declares.
• MANDATORY Disjoint write sets per wave · all-return barrier before the next wave · specialist routing · sub-agents NEVER fan out further unless their own agent definition authorizes it.

Project Protocol Overlay — Before executing this skill, resolve any PROJECT overlay rules layered onto it: match this skill's name against the Target column of the project's skill-protocol index (docs/project-reference/skill-protocols-reference.md by default; a referenceDocs entry in docs/project-config.json overrides the path), taking the most specific matching tier ONLY — exact name > glob > *. That precedence orders overlays against EACH OTHER, never against this skill. Read ONLY the matched bodies, resolved as <protocols-dir>/<Name>.md; a row's Body link is display text, never a read path. A matched body that is missing or malformed is REPORTED and skipped — never reconstructed from the index Description. No index, or no match -> proceed with no overlay, silently. Full contract: .claude/skills/project-skill-protocol/references/registry.md.

Overlays are ADDITIVE ONLY: they ADD rules on top of this skill's own protocol and NEVER replace, override, disable, or reinterpret a rule it already states — removing every overlay must return this skill to exactly its documented behavior. An overlay is a BRIEF, not an authority escalation: it can NEVER waive a workflow gate, git discipline, a review gate, or a user-confirmation gate. A genuine overlay-vs-skill conflict, or two equally-specific overlays that directly contradict -> surface both to the user; NEVER resolve silently.

MUST ATTENTION resolve project protocol overlays for this skill BEFORE executing — most specific matching tier only (exact > glob > *, which ranks overlays against each other, NEVER against this skill), read only matched bodies at <protocols-dir>/<Name>.md; a missing or malformed body is reported, never reconstructed. Overlays are ADDITIVE ONLY (they never replace this skill's own rules) and are a brief, NEVER an authority escalation; an equal-specificity contradiction goes to the user.

Closing Reminders

IMPORTANT MUST ATTENTION Goal: any agent, with or without hooks, reaches a verified project-context state before project-specific work.

MUST ATTENTION Protocols in force (concise digest of the SYNC/shared blocks this skill carries):

• Critical Thinking: ALWAYS apply critical + sequential thinking; traced proof, confidence >80% to act.
• AI Mistake Prevention: verify generated content against evidence, trace downstream references, verify all affected outputs, re-read after context loss, surface ambiguity.
• Parallel Sub-Agent Dispatch: Tag tasks PAR/SEQ, group PAR into disjoint-write-set waves, spawn each wave in ONE message, barrier before advancing.

IMPORTANT MUST ATTENTION use $project-init as the unified missing-context route; lower-level skills remain implementation steps. IMPORTANT MUST ATTENTION before doing anything, create many small task-plan rows covering every setup phase, the post-config parallel group, and the final review skill calls. IMPORTANT MUST ATTENTION after config exists, call $scan-all and $workflow-code-to-spec as a parallel group when possible, then wait for both outcomes. IMPORTANT MUST ATTENTION never finish $project-init without Call $scan-all and Call $workflow-code-to-spec after config initialization, followed by final rows Call $changes-review, Call $why-review, Spawn background $graph-build sub-agent. IMPORTANT MUST ATTENTION final $graph-build runs in a background sub-agent after setup/review/verification is otherwise done; track it to returned outcome or explicit blocker. IMPORTANT MUST ATTENTION for content-bearing projects, invoke $workflow-code-to-spec; do not close on a recommendation/handoff alone. IMPORTANT MUST ATTENTION record explicit blocker for any unavailable required skill/tool; silent skip is not completion. IMPORTANT MUST ATTENTION preserve user-authored root instruction files; do not overwrite project-only content. IMPORTANT MUST ATTENTION rerun Phase 0 after every setup phase because the next route depends on current evidence. IMPORTANT MUST ATTENTION keep reusable setup logic project-neutral; project-specific facts belong in config/reference docs.

Anti-Rationalization:

Critical Thinking Mindset — Apply critical thinking, sequential thinking. Every claim needs traced proof, confidence >80% to act. Anti-hallucination: Never present guess as fact — cite sources for every claim, admit uncertainty freely, self-check output for errors, cross-reference independently, stay skeptical of own confidence — certainty without evidence root of all hallucination.

AI Mistake Prevention — Failure modes to avoid on every task:

Re-read files after context changes. Context compaction, resume, or long-running work can make memory stale; verify current files before acting. Verify generated content against source evidence. AI hallucinates APIs, names, claims, and document facts. Check the relevant source before documenting or referencing. Check downstream references before deleting or renaming. Removing an artifact can stale docs, generated mirrors, configs, and callers; map references first. Trace the full impact chain after edits. Changing a definition can miss derived outputs and consumers. Follow the affected chain before declaring done. Verify ALL affected outputs, not just the first. One green check is not all green checks; validate every output surface the change can affect. Assume existing values are intentional — ask WHY before changing OR flagging one as a defect. Before changing or reporting a constant, limit, flag, cutoff, wording, or pattern, read nearby context and history, the CALLER's ordering, and 2+ sibling call sites of the same convention. A doc stating WHAT without WHY is missing rationale, not proof of a missing guard. Surface ambiguity before acting — don't pick silently. Multiple valid interpretations require an explicit question or stated assumption with risk. Assert the outcome your system owns, not the intermediate state your infrastructure owns. When verifying async work, assert the final business state — never the delivery/retry bookkeeping held in shared infrastructure that any co-running process can write. Such a check passes when run alone and flakes the moment anything else shares that infrastructure. Keep shared guidance role-relevant. Universal guidance must help every receiving skill or agent; code-specific obligations belong only in code-specific protocols.

MUST ATTENTION apply critical + sequential thinking — every claim needs appropriate traced evidence (file:line for repo/code claims; source URL or artifact section for research, product, content, and docs claims); confidence >80% to act, <60% DO NOT recommend. Anti-hallucination: never present guess as fact, admit uncertainty freely, cross-reference independently, stay skeptical of own confidence.

MUST ATTENTION apply AI mistake prevention — verify generated content against evidence, trace downstream references before deleting or renaming, verify all affected outputs, re-read files after context loss, and surface ambiguity before acting.

Hookless Prompt Protocol Mirror (Auto-Synced)

Source: .claude/.ck.json + .claude/skills/shared/sync-inline-versions.md (:full blocks) + .claude/scripts/lib/hookless-prompt-protocol.cjs

[WORKFLOW-EXECUTION-PROTOCOL] [BLOCKING] Workflow Execution Protocol — MANDATORY IMPORTANT MUST CRITICAL. Do not skip for any reason.

Generic portability boundary: Reusable skills and protocol text stay project-neutral; project-specific conventions are discovered from docs/project-config.json and docs/project-reference/. Apply shared AI-SDD from shared/sdd-artifact-contract.md. Read docs/project-config.json and docs/project-reference/docs-index-reference.md, then open the project reference docs named there. For spec, test-case, behavior-change, public-contract, or docs/specs/ work, route through the local spec docs named by the docs index: feature-spec-reference.md, spec-system-reference.md, spec-principles.md, and workflow-spec-test-code-cycle-reference.md when specs/tests/code must stay synchronized. If either file or a required reference doc is missing or stale, auto-run $project-init (or the narrow lower-level route such as $project-config, $docs-init, $scan-all, or $scan --target=<key>) before ordinary project-specific work. Any supported AI tool may execute when this shared context and local docs are available.

1. DETECT: If the prompt starts with an explicit slash skill/workflow command, execute it directly. Otherwise match the prompt against the workflow catalog and skill list.
2. ANALYZE: Choose the best option: execute directly, invoke a skill, activate a standard workflow, or compose a custom step combination.
3. AUTO-SELECT: Pick the best option yourself. Do not ask the user to choose between direct execution, skill, standard workflow, or custom workflow.
4. ACTIVATE: For a selected workflow, call $start-workflow <workflowId>; for a selected skill, invoke that skill; for a custom workflow, sequence custom steps directly; for direct execution, proceed with the task.
5. CREATE TASKS: task tracking for ALL workflow/skill/custom steps before execution when the selected path has multiple steps.
6. PARALLELIZE: Before executing the task list, tag each task PAR (independent inputs + write set disjoint from every other PAR task) or SEQ (name the blocking dependency), group PAR tasks into waves, declare the wave plan, and spawn each wave's sub-agents in ONE message — all-return barrier per wave, fan-out one level deep unless a sub-agent's own definition authorizes further fan-out. Sequential-by-default is a defect when tasks are independent; do not parallelize shared write targets, output-consuming tasks, trivial single-file work, ordering a skill or workflow explicitly fixes, or user-approval gates.
7. EXECUTE: Advance per the Workflow Step Advancement & Parallel Phases rule in your context instructions — model-driven; a sub-agent completion advances a step identically to an inline call; a parallel-phase group is an all-return barrier (advance only after ALL members return, never serialize it)

Shared AI-SDD Protocol Markers

Source: .claude/skills/shared/sync-inline-versions.md

SYNC:ai-sdd-artifact-contract

AI-SDD Artifact Contract — Shared spec-driven development rules stay portable and source-owned.

1. Keep reusable AI-SDD principles in .claude; put repository-specific paths, commands, owners, products, and formats in project config/reference docs.
2. Preserve cycle: spec -> plan -> tasks -> implement -> verify -> update spec/docs.
3. Trace every requirement or invariant through decision, task, TC/test, source evidence, and docs/spec update.
4. Treat code-to-spec extraction as reference-only until accepted by the canonical spec owner.
5. Any supported AI tool may plan, implement, review, or verify with synced context; using multiple tools is optional.
6. Update .claude source first, then sync generated mirrors; do not manually edit .agents, .codex, or AGENTS.md. — why: mirrors are generated artifacts; hand-edits are overwritten on the next sync
7. If docs/project-config.json, root instruction files, or a required project-reference doc is missing or stale, auto-run $project-init or the narrow lower-level route before ordinary project-specific work.

Active reference: shared/sdd-artifact-contract.md in the active skills root.

SYNC:ai-sdd-artifact-contract:reminder

• MANDATORY Apply shared/sdd-artifact-contract.md; keep reusable AI-SDD in .claude and local rules in project docs.
• MANDATORY Code-to-spec extraction is reference-only until canonical acceptance; any supported AI tool may execute with synced context.
• MANDATORY Update .claude source before syncing generated mirrors; do not manually edit .agents, .codex, or AGENTS.md.
• MANDATORY Missing or stale project config, root instruction files, or required reference docs route project-specific work through $project-init or the narrow setup route automatically. [TASK-PLANNING] [MANDATORY] BEFORE executing any workflow or skill step, create/update task tracking for all planned steps, then keep it synchronized as each step starts/completes.

[LESSON-LEARNED-REMINDER] [BLOCKING] Task Planning & Continuous Improvement — MANDATORY. Do not skip.

Break work into small tasks (task tracking) before starting. Add final task: "Analyze AI mistakes & lessons learned".

Extract lessons — ROOT CAUSE ONLY, not symptom fixes:

1. Name the FAILURE MODE (reasoning/assumption failure), not symptom — "assumed API existed without reading source" not "used wrong enum value".
2. Generality test: does this failure mode apply to ≥3 contexts/codebases? If not, abstract one level up.
3. Write as a universal rule — strip project-specific names/paths/classes. Useful on any codebase.
4. Consolidate: multiple mistakes sharing one failure mode → ONE lesson.
5. Recurrence gate: "Would this recur in future session WITHOUT this reminder?" — No → skip $learn.
6. Auto-fix gate: "Could $code-review/$code-simplifier/$security-review/$lint catch this?" — Yes → improve review skill instead.
7. BOTH gates pass → ask user to run $learn. [CRITICAL-THINKING-MINDSET] Apply critical thinking, sequential thinking. Every claim needs traced proof, confidence >80% to act. Anti-hallucination principle: Never present guess as fact — cite sources for every claim, admit uncertainty freely, self-check output for errors, cross-reference independently, stay skeptical of own confidence — certainty without evidence root of all hallucination. AI Attention principle (Primacy-Recency): Put the 3 most critical rules at both top and bottom of long prompts/protocols so instruction adherence survives long context windows. Goal-driven execution: Define success criteria first, loop until verified, and stop only when observable checks pass. Tests verify intent: Tests must protect business rules/invariants and fail when the protected intent breaks, not only mirror current behavior.

Common AI Mistake Prevention (System Lessons)

• Re-read files after context compaction. Edit requires prior Read in same context; compaction wipes read state. Re-read before editing.
• Grep for old terms after bulk replacements. AI over-trusts find/replace completeness. Grep full repo after bulk edits for missed refs in docs/configs/catalogs.
• Check downstream references before deleting. Deletions cascade doc/code staleness. Map referencing files before removal.
• After memory loss, check existing state before creating new. Compaction wipes prior-work memory. Query current state to resume — never blindly duplicate.
• Verify AI-generated content against actual code. AI hallucinates APIs, class names, method signatures. Grep to confirm existence before documenting/referencing.
• Trace full dependency chain after edits. Changing a definition misses downstream consumers. Trace the full chain.
• When renaming, grep ALL consumer file types. Some file types silently ignore missing refs (no compile error). Search code, templates, configs, generated files.
• Trace ALL code paths when verifying correctness. Code existing ≠ code executing. Trace early exits, error branches, conditional skips — not just happy path.
• Update docs that embed canonical data when source changes. Docs inlining derived data (workflows, schemas, configs) go stale silently. Update all embedding docs alongside source.
• Verify sub-agent results after context recovery. Background agents may finish while parent compacted — grep-verify output, don't trust assumed completion.
• Cross-check full target list against sub-agent assignments. Parallel sub-agents by category miss boundary items. Reconcile union of assignments against target list before proceeding.
• Sub-agents inherit knowledge only from their agent .md definition — use custom agent types, not built-in Explore. Tool adoption = permission + knowledge + enforcement (numbered workflow step).
• Persist sub-agent findings incrementally, not as a final batch. Long sub-agents hit cutoffs before final write — findings lost. Instruct append-per-section to report file.
• When debugging, ask "whose responsibility?" before fixing. Trace caller (wrong data) vs callee (wrong handling). Fix at responsible layer — never patch symptom site.
• Test failure → record a provisional verdict before trace/edit, then investigate. Use the full five-way taxonomy: SOURCE-WRONG (production violates intent), TEST-WRONG (assertion/setup is stale), TEST-NOT-OPTIMAL (valid but fragile or low-signal test), ENVIRONMENT-BLOCKED (external state prevents a verdict), or AMBIGUOUS (intent/evidence cannot choose safely). Then trace root cause and triangulate against the governing spec (docs/specs/** if one exists) AND source. NEVER weaken an assertion, add a skip, relax a timeout, or change source merely to force green.
• Grep ALL removed names after extraction/refactoring. Primary file "done" ≠ secondary files clean. Grep entire scope for every removed symbol before declaring complete.
• Assume existing values are intentional — ask WHY before changing OR flagging one as a defect. Pattern-matching as "wrong" skips context. Before changing or reporting any constant/limit/flag/cutoff: read comments, git blame, the CALLER's ordering (the guarantee that makes the value correct usually lives in code running immediately BEFORE the cited line), and 2+ sibling call sites of the same convention. A doc stating WHAT without WHY is missing rationale, not proof of a missing guard — and in a validation pass, an accurate file:line citation proves the transcription, never the defect.
• Verify ALL affected outputs, not just the first. One build green ≠ all green. Multi-stack changes (backend/frontend/tests/docs) require verifying EVERY output.
• Evaluate fit before copying a nearby pattern. Closest example ≠ matching preconditions — verify the new context shares the same constraints, base classes, scope, lifetime.
• Holistic-first debugging — resist nearest-attention trap. Don't dive into first plausible cause. List EVERY precondition (config, env vars, paths, DB, endpoints, creds, versions, DI, data). Verify each against evidence (grep/query — not reasoning). Ask "what would falsify this?" — if nothing, it's not a hypothesis. Most expensive failure: going deeper in "obvious" layer while bug sits in layer never questioned.
• Surgical changes — apply the diff test (context-aware). Two modes: (1) Bug fix → every line traces to the bug; no restyling; orphan cleanup only for imports YOUR changes made unused. (2) Review/enhancement → implement improvements AND announce as "Enhancement beyond main request: [what]". Never silently scope-creep. Diff test: "Would this line exist if I wasn't asked to do X?" — if no, delete or announce.
• Surface ambiguity before coding — don't pick silently. Multiple valid interpretations → present each with effort: "[Request] could mean (1) [N h], (2) [N h]. Which matters?" List scope/format/volume/constraints assumptions first. If simpler path exists, say so. Never silently pick.
• [MANDATORY FIRST ACTION] ALWAYS activate a suitable skill or workflow BEFORE responding. Match task against workflow catalog + skill list; invoke via skill invocation or $start-workflow <workflowId>. NEVER answer or write code before checking. Skip = protocol violation.
• Why-Review adversarial mindset — apply when reviewing any plan, decision, or design. Default SKEPTIC not VALIDATOR: steel-man a rejected alternative, invert each stated reason ("what does it sacrifice?"), stress-test top 2-3 assumptions, run pre-mortem ("ships, fails in 3 months — what breaks?"), surface 1-2 alternatives author missed. Section presence ≠ quality; quality = causal reasoning + concrete mitigations + evidence, not "it's better" or "monitor closely".
• Front-load report-write in sub-agent prompts for large reviews. Many-file sub-agents hit budget before final write — findings lost. Design prompts so: (1) report-write is first explicit deliverable, (2) append per-file/section (not batched), (3) scope bounded so reads don't exhaust budget. Truncated mid-sentence with no report file → spawn narrower scope, don't retry same prompt.
• After context compaction, re-verify all prior phase outcomes before continuing. Summaries describe intent, not environment state (git index, filesystem, processes). On resume, FIRST audit: git status, re-read modified files, verify filesystem. Every "completed" claim is an untested hypothesis until evidence confirms.
• OOM/memory: check row count before row size. Triage: (1) Unbounded query — no DB filter for trigger? Push filter to DB; eliminates OOM. (2) Large rows? Projection reduces proportionally. Row reduction > projection in ROI.
• Assert the outcome your system OWNS, never the intermediate state your INFRASTRUCTURE owns. When testing anything asynchronous (queue/broker delivery, retries, background jobs, caches, replication), assert the final business/entity state. NEVER assert the delivery bookkeeping — consume/send status, attempt counts, last-error, row existence or counts in a broker, scheduler, or outbox/inbox table. That bookkeeping lives in shared infrastructure that ANY co-running process (a peer worker, a second replica, a leftover local container) can write, usually under a deterministic shared key, so the assertion silently tests the developer's environment instead of the system: green when run alone, flaky the instant anything else shares that broker + database. Gate question for every assertion: "would this hold no matter WHICH process did the work?" — if no, assert the converged data state instead. Corollary: process-local fault injection and in-process telemetry cannot gate work any process may perform — use them as stress amplifiers (arm → bounded window → disarm → assert convergence), never as preconditions.
• Keep domain concepts out of generic/shared/infrastructure layers. Reusable layer (shared library, framework, infra module) must reference NO consumer-specific domain concept — tenant/customer/product IDs, business entities, feature rules. Leak compiles + runs → passes review silently while coupling the "reusable" layer to one consumer. Keep shared type domain-free; push domain fields/logic down into the consumer via subclass/composition. — why: a layer coupled to one consumer's domain is no longer reusable.