@windyroad/itil 1.2.0 → 2.0.0-preview.1111

package/hooks/lib/changeset-detect.sh

@@ -1,12 +1,12 @@
 #!/bin/bash
-# P141: shared changeset-discipline detection helper.
+# afk-iter-packages-plugin-commits-without-changesets-orchestrator-main-turn-back-fill-is-fragile-recovery-hook-level-enforcement-preferable-problem: shared changeset-discipline detection helper.
 #
 # `detect_changeset_required` returns 0 (no change required — allow) /
 # 1 (changeset required but not staged — caller should deny). On 1, the
 # offending plugin slug is echoed on stdout so callers can name it in
 # deny messages without re-parsing diff output.
 #
-# Trap shape (P141):
+# Trap shape (afk-iter-packages-plugin-commits-without-changesets-orchestrator-main-turn-back-fill-is-fragile-recovery-hook-level-enforcement-preferable-problem):
 #   `/wr-itil:work-problems` AFK iter subprocesses receive prompt-time
 #   guidance to author a `.changeset/*.md` whenever they ship a
 #   `packages/<plugin>/` change. Under context pressure (heavy SKILL.md
@@ -26,7 +26,7 @@
 #           - allow-list: `README.md`.
 #           - allow-list: `docs/<anything>.md` (per architect verdict
 #             2026-05-02 — `*.md` under `docs/` only; SKILL.md is the
-#             publishable contract per ADR-037 framing and is NOT in
+#             publishable contract per skill-testing-strategy-contract-assertion-bats-companion-to-architecture-rule framing and is NOT in
 #             the allow-list).
 #           - otherwise: publishable source — record the slug.
 #       * any other path: ignored (non-publishable surface — `.github/`,
@@ -44,8 +44,8 @@
 #         is required.
 #       * Neither check satisfied → return 1 + echo the slug.
 #
-# Phase 2 rationale (P141 2026-05-31): AFK orchestrator iters that
-# ship a multi-commit slice for one plugin (e.g. P346 Phase 3 across
+# Phase 2 rationale (afk-iter-packages-plugin-commits-without-changesets-orchestrator-main-turn-back-fill-is-fragile-recovery-hook-level-enforcement-preferable-problem 2026-05-31): AFK orchestrator iters that
+# ship a multi-commit slice for one plugin (e.g. wr-itil-review-problems-has-no-path-to-close-tickets-that-are-no-longer-relevant-evidence-based-not-age-based-structural-outflow-gap-drives-monotonic-backlog-growth-problem Phase 3 across
 # 4 commits, 2 of which touched `packages/itil/`) should not author N
 # redundant changesets for one logical bump. changesets-action
 # collapses bump-class at version-package time, so per-commit
@@ -72,32 +72,32 @@
 # Cost: one `git diff` invocation per check (~10ms on this repo's
 # working tree). Per-invocation deterministic — runs on every
 # `git commit` invocation rather than relying on per-tool-call
-# session state tracking. Mirrors the P125 `staging-detect.sh`
+# session state tracking. Mirrors the staging-trap-recurs-despite-documentation-hook-level-enforcement-candidate-problem `staging-detect.sh`
 # precedent (architect-approved no-marker design).
 #
 # References:
-#   ADR-005  — plugin testing strategy (hook bats live under
-#              `hooks/test/` per P081 behavioural-test discipline).
-#   ADR-013 Rule 1 — deny redirects with mechanical recovery (the deny
+#   plugin-testing-strategy-architecture-rule  — plugin testing strategy (hook bats live under
+#              `hooks/test/` per problem-081-structural-source-content-tests-are-wasteful-tdd-agent-should-reject-them-and-require-behavioural-tests-framework-stub-enhancements-problem behavioural-test discipline).
+#   structured-user-interaction-for-governance-skill-decisions-architecture-rule Rule 1 — deny redirects with mechanical recovery (the deny
 #              text names the plugin slug + the literal `bun run
 #              changeset` command + the BYPASS env var override).
-#   ADR-014  — governance skills commit their own work (this hook
+#   governance-skills-commit-their-own-completed-work-architecture-rule  — governance skills commit their own work (this hook
 #              ensures iter commits stay self-contained per
-#              ADR-014 single-commit grain).
-#   ADR-018  — inter-iteration release cadence (this hook strengthens
+#              governance-skills-commit-their-own-completed-work-architecture-rule single-commit grain).
+#   inter-iteration-release-cadence-for-afk-loops-architecture-rule  — inter-iteration release cadence (this hook strengthens
 #              the cadence by ensuring every publishable iter has a
 #              changeset to drain at release time).
-#   ADR-038  — progressive disclosure / deny-message terseness.
-#   ADR-045  — hook injection budget (Pattern 1 silent-on-pass; deny
+#   progressive-disclosure-once-per-session-budget-for-userpromptsubmit-governance-prose-architecture-rule  — progressive disclosure / deny-message terseness.
+#   hook-injection-budget-policy-for-pretooluse-and-posttooluse-hooks-architecture-rule  — hook injection budget (Pattern 1 silent-on-pass; deny
 #              band ≤300 bytes for this hook).
-#   P073     — sibling changeset author-time gate (different surface:
+#   problem-073-no-voice-and-tone-check-or-risk-assessment-on-changeset-bodies-which-populate-changelog-md-release-prs-github-releases-and-npm-release-notes-problem     — sibling changeset author-time gate (different surface:
 #              Write/Edit on `.changeset/*.md`). Composes-with as
 #              defence-in-depth.
-#   P125     — sibling staging-trap helper (same enforcement-layer
+#   staging-trap-recurs-despite-documentation-hook-level-enforcement-candidate-problem     — sibling staging-trap helper (same enforcement-layer
 #              shape — per-invocation deterministic, no markers).
-#   P141     — this helper.
+#   afk-iter-packages-plugin-commits-without-changesets-orchestrator-main-turn-back-fill-is-fragile-recovery-hook-level-enforcement-preferable-problem     — this helper.
 
-# P141 Phase 2 helper — does any `.changeset/*.md` ALREADY in scope target the plugin
+# afk-iter-packages-plugin-commits-without-changesets-orchestrator-main-turn-back-fill-is-fragile-recovery-hook-level-enforcement-preferable-problem Phase 2 helper — does any `.changeset/*.md` ALREADY in scope target the plugin
 # slug via its YAML frontmatter `"@windyroad/<slug>": <bump>` line?
 #
 # Scope = files reachable from HEAD but not from `origin/<base>`,
@@ -170,7 +170,7 @@ EOF
   return 0
 }
 
-# P387 helper — echo the space-separated, upper-cased, de-duplicated set of
+# changeset-discipline-check-2b-is-plugin-scoped-not-change-scoped-plugin-source-commits-ship-undocumented-when-a-sibling-changeset-already-targets-the-plugin-problem helper — echo the space-separated, upper-cased, de-duplicated set of
 # work-item IDs found in the text passed as $1. Work-item identity = problem
 # ticket (`P<NNN>`), RFC (`RFC-<NNN>`), or story (`STORY-<NNN>`). ADR refs are
 # deliberately excluded — an ADR is cross-cutting context cited in passing, not
@@ -193,9 +193,9 @@ _work_item_ids() {
 # not satisfied by either staged Check 2a or in-scope Check 2b.
 #
 # $1 (optional) — the git-commit COMMAND string (or commit message). Its
-#   work-item ID(s) are matched against in-scope changesets for the P387
+#   work-item ID(s) are matched against in-scope changesets for the changeset-discipline-check-2b-is-plugin-scoped-not-change-scoped-plugin-source-commits-ship-undocumented-when-a-sibling-changeset-already-targets-the-plugin-problem
 #   change-scoped Check 2b. Empty / omitted → Check 2b falls back to the
-#   pre-P387 plugin-scoped behaviour (any covering changeset allows), which
+#   pre-changeset-discipline-check-2b-is-plugin-scoped-not-change-scoped-plugin-source-commits-ship-undocumented-when-a-sibling-changeset-already-targets-the-plugin-problem plugin-scoped behaviour (any covering changeset allows), which
 #   is the conservative choice when no commit context is available.
 #
 # Echoes the offending plugin slug on stdout when detected.
@@ -284,27 +284,27 @@ EOF
     return 0
   fi
 
-  # Check 2b (P141 Phase 2 + P387 change-scoped) — an in-scope changeset
+  # Check 2b (afk-iter-packages-plugin-commits-without-changesets-orchestrator-main-turn-back-fill-is-fragile-recovery-hook-level-enforcement-preferable-problem Phase 2 + changeset-discipline-check-2b-is-plugin-scoped-not-change-scoped-plugin-source-commits-ship-undocumented-when-a-sibling-changeset-already-targets-the-plugin-problem change-scoped) — an in-scope changeset
   # targeting the plugin satisfies, but only when it is change-scoped to
   # THIS commit. Scope = unpushed-range commits + untracked + modified-
   # not-staged working-tree files. Once consumed onto origin, the changeset
   # is gone and a fresh one is required.
   local covering
   if covering=$(_changeset_in_scope_covers_plugin "$plugin_source_slug"); then
-    # P387: tighten plugin-scoped → change-scoped. A plugin can carry a
-    # changeset for an unrelated change; before P387 that wrongly covered
+    # changeset-discipline-check-2b-is-plugin-scoped-not-change-scoped-plugin-source-commits-ship-undocumented-when-a-sibling-changeset-already-targets-the-plugin-problem: tighten plugin-scoped → change-scoped. A plugin can carry a
+    # changeset for an unrelated change; before changeset-discipline-check-2b-is-plugin-scoped-not-change-scoped-plugin-source-commits-ship-undocumented-when-a-sibling-changeset-already-targets-the-plugin-problem that wrongly covered
     # THIS commit, shipping it to npm with no CHANGELOG record of its own
-    # (witnessed: P164's fix rode P374's changeset). Deny only on positive
+    # (witnessed: latent-octal-eval-bug-in-next-id-formula-across-all-4-ticket-creator-skills-local-max-1-fails-with-value-too-great-for-base-when-local-max-reaches-099-problem's fix rode extract-risks-from-reports-sh-hardcodes-windy-road-agent-plugins-suite-branding-in-the-adopter-generated-docs-risks-readme-md-problem's changeset). Deny only on positive
     # evidence the covering changeset(s) belong to a DIFFERENT change: the
     # commit cites work-item ID(s), EVERY covering changeset cites work-item
     # ID(s), and none overlap. Any ambiguity allows — a ticket-less commit,
-    # a prose-only changeset, or an ID overlap — so the ADR-014 batch-grain
+    # a prose-only changeset, or an ID overlap — so the governance-skills-commit-their-own-completed-work-architecture-rule batch-grain
     # (same-slice commits share a ticket) and prose-only / adopter changesets
     # are never over-fired.
     local commit_ids cs_path cs_ids id has_idless=0 overlap=0
     commit_ids=$(_work_item_ids "$commit_msg")
 
-    # Commit cites no work-item ID → cannot change-scope; allow (pre-P387).
+    # Commit cites no work-item ID → cannot change-scope; allow (pre-changeset-discipline-check-2b-is-plugin-scoped-not-change-scoped-plugin-source-commits-ship-undocumented-when-a-sibling-changeset-already-targets-the-plugin-problem).
     [ -n "$commit_ids" ] || return 0
 
     while IFS= read -r cs_path; do

package/hooks/lib/claude-space-gate.sh

@@ -1,5 +1,5 @@
 #!/bin/bash
-# Shared gate logic for `.claude/` user-space write protection (P131).
+# Shared gate logic for `.claude/` user-space write protection (agents-write-project-generated-artefacts-under-claude-user-controlled-config-space-gate-exclusions-are-read-tolerance-not-write-permission-problem).
 #
 # Sourced by itil-claude-space-protection.sh. Provides:
 #   is_protected_claude_path  — returns 0 if path is project-scoped `.claude/`
@@ -14,12 +14,12 @@
 # This gate enforces a PERSISTENT "user pre-approved this specific path"
 # marker — different lifecycle (no TTL, no drift, never auto-cleared) and
 # different scope (file-path-keyed, not session-keyed). Architect verdict
-# (P131 Phase 2): keep semantically distinct from existing gate libraries.
+# (agents-write-project-generated-artefacts-under-claude-user-controlled-config-space-gate-exclusions-are-read-tolerance-not-write-permission-problem Phase 2): keep semantically distinct from existing gate libraries.
 #
 # Marker convention: `.claude/.agent-write-approved-<sha256-of-rel-path>`
 # where `<sha256-of-rel-path>` is the SHA-256 hex of the path relative to
 # project root (PWD). Persistent, in-tree, file-existence test — same
-# shape as ADR-030 / P120 `.claude/.install-updates-consent` precedent.
+# shape as repo-local-skills-for-project-specific-workflow-tooling-architecture-rule / install-updates-consent-gate-fires-on-every-invocation-despite-the-answer-being-invariably-all-confirmed-siblings-problem `.claude/.install-updates-consent` precedent.
 #
 # Allow-list scope: paths under `.claude/` that are user-controlled by
 # convention (Claude Code config, user-authored extensions, system state):
@@ -30,18 +30,18 @@
 #   - .agent-write-approved-* markers themselves (so user can create them)
 #
 # References:
-#   ADR-009 — gate marker lifecycle (this gate adds a NEW persistent
+#   gate-marker-lifecycle-ttl-drift-not-stop-hook-reset-architecture-rule — gate marker lifecycle (this gate adds a NEW persistent
 #             marker class; the ADR's session-scoped /tmp markers are
 #             unchanged)
-#   ADR-013 Rule 6 — non-interactive fail-safe (parse-error => exit 0)
-#   ADR-030 — install-updates persistent in-tree consent marker precedent
-#   ADR-038 — progressive disclosure (deny message <500 bytes)
-#   ADR-045 — hook injection budget (Pattern 1 silent-on-pass; allow path
+#   structured-user-interaction-for-governance-skill-decisions-architecture-rule Rule 6 — non-interactive fail-safe (parse-error => exit 0)
+#   repo-local-skills-for-project-specific-workflow-tooling-architecture-rule — install-updates persistent in-tree consent marker precedent
+#   progressive-disclosure-once-per-session-budget-for-userpromptsubmit-governance-prose-architecture-rule — progressive disclosure (deny message <500 bytes)
+#   hook-injection-budget-policy-for-pretooluse-and-posttooluse-hooks-architecture-rule — hook injection budget (Pattern 1 silent-on-pass; allow path
 #             emits zero bytes)
-#   P119    — manage-problem-enforce-create.sh precedent for itil
+#   agent-bypasses-wr-itil-manage-problem-step-2-duplicate-check-by-writing-tickets-directly-to-docs-problems-problem    — manage-problem-enforce-create.sh precedent for itil
 #             PreToolUse:Write enforcement hook shape
-#   P120    — install-updates consent-marker persistence precedent
-#   P131    — this gate's driver
+#   install-updates-consent-gate-fires-on-every-invocation-despite-the-answer-being-invariably-all-confirmed-siblings-problem    — install-updates consent-marker persistence precedent
+#   agents-write-project-generated-artefacts-under-claude-user-controlled-config-space-gate-exclusions-are-read-tolerance-not-write-permission-problem    — this gate's driver
 
 # Returns 0 (true) if FILE_PATH is a project-scoped .claude/ path AND is
 # NOT in the user-space allow-list. Returns 1 (false) otherwise.
@@ -79,9 +79,9 @@ is_protected_claude_path() {
   # Strip leading ./ if present
   rel_path="${rel_path#./}"
 
-  # Must be under .claude/ at project-relative root depth.
+  # Must be under a runtime config directory at project-relative root depth.
   case "$rel_path" in
-    .claude/*) ;;
+    .claude/*|.codex/*) ;;
     *) return 1 ;;
   esac
 
@@ -114,6 +114,13 @@ is_protected_claude_path() {
     # Claude Code's own state: projects/<id>/memory/, worktrees/
     .claude/projects/*) return 1 ;;
     .claude/worktrees/*) return 1 ;;
+    # Codex runtime configuration and generated runtime surfaces.
+    .codex/config.toml) return 1 ;;
+    .codex/AGENTS.md) return 1 ;;
+    .codex/hooks.json) return 1 ;;
+    .codex/agents/*) return 1 ;;
+    .codex/skills/*) return 1 ;;
+    .codex/.agent-write-approved-*) return 1 ;;
   esac
 
   # Project-scoped .claude/ path NOT on the allow-list — protected.
@@ -156,11 +163,12 @@ has_approval_marker() {
   fi
 
   [ -n "$hash" ] || return 1
-  [ -f "${pwd_root}/.claude/.agent-write-approved-${hash}" ]
+  local config_dir="${rel_path%%/*}"
+  [ -f "${pwd_root}/${config_dir}/.agent-write-approved-${hash}" ]
 }
 
 # Emit fail-closed PreToolUse deny JSON. Reason should be terse (<500
-# bytes per ADR-038). Hook callers compose the basename + reason inline.
+# bytes per progressive-disclosure-once-per-session-budget-for-userpromptsubmit-governance-prose-architecture-rule). Hook callers compose the basename + reason inline.
 #
 # Usage: claude_space_deny "BLOCKED: <reason>"
 claude_space_deny() {

package/hooks/lib/command-detect.sh

@@ -1,5 +1,5 @@
 #!/bin/bash
-# P268: shared command-detection helper for PreToolUse:Bash hooks
+# hook-substring-matches-git-commit-anywhere-in-bash-command-not-just-actual-git-commit-invocations-problem: shared command-detection helper for PreToolUse:Bash hooks
 # that need to distinguish ACTUAL `git commit` invocations from Bash
 # commands that merely MENTION the literal phrase "git commit" in
 # argument vectors or heredoc bodies (grep patterns, sed patterns,
@@ -14,11 +14,11 @@
 #
 # that 5 sibling PreToolUse:Bash hooks previously used to gate on
 # `git commit`. The substring match misfired on legitimate non-commit
-# Bash whose arguments contained the phrase (P268 ticket Description
+# Bash whose arguments contained the phrase (hook-substring-matches-git-commit-anywhere-in-bash-command-not-just-actual-git-commit-invocations-problem ticket Description
 # — observed iter-1 retro write and orchestrator grep, ≥3 events per
 # session).
 #
-# Strategy (Fix shape B per P268 ticket):
+# Strategy (Fix shape B per hook-substring-matches-git-commit-anywhere-in-bash-command-not-just-actual-git-commit-invocations-problem ticket):
 #   1. Strip leading whitespace from the candidate command string.
 #   2. Iteratively strip env-var-assignment prefixes (`VAR=value `,
 #      `VAR="..." `, `VAR='...' `) and a `cd <path> &&` prefix until
@@ -28,7 +28,7 @@
 #      string. The end-of-token boundary check stops `git commit-tree`
 #      and similar `git commit-*` plumbing commands from matching.
 #
-# Scope deliberately narrow (per P268 ticket Description recommended
+# Scope deliberately narrow (per hook-substring-matches-git-commit-anywhere-in-bash-command-not-just-actual-git-commit-invocations-problem ticket Description recommended
 # Fix shape B): handles only the prefix shapes the orchestrator and
 # capture/manage/work skills actually emit — direct `git commit`,
 # `cd && git commit`, `VAR=value git commit`. Does NOT split on
@@ -46,20 +46,20 @@
 # `lib/readme-refresh-detect.sh` echoes the offending ticket path).
 #
 # References:
-#   ADR-005  — plugin testing strategy (this helper's bats live at
+#   plugin-testing-strategy-architecture-rule  — plugin testing strategy (this helper's bats live at
 #              `packages/itil/hooks/test/command-detect.bats` per
 #              behavioural-test discipline).
-#   ADR-009  — gate marker lifecycle (helper is per-invocation, no
+#   gate-marker-lifecycle-ttl-drift-not-stop-hook-reset-architecture-rule  — gate marker lifecycle (helper is per-invocation, no
 #              marker state — same precedent as `lib/staging-detect.sh`,
 #              `lib/changeset-detect.sh`, `lib/readme-refresh-detect.sh`).
-#   ADR-045  — hook injection budget (silent-on-pass / silent-on-fail
+#   hook-injection-budget-policy-for-pretooluse-and-posttooluse-hooks-architecture-rule  — hook injection budget (silent-on-pass / silent-on-fail
 #              — pure exit-code contract).
-#   P125     — sibling staging-trap hook (candidate for follow-up
+#   staging-trap-recurs-despite-documentation-hook-level-enforcement-candidate-problem     — sibling staging-trap hook (candidate for follow-up
 #              refactor to use this helper).
-#   P141     — sibling changeset-discipline hook (candidate).
-#   P165     — sibling README-refresh-discipline hook (first consumer
-#              under P268).
-#   P268     — this helper's surfacing ticket.
+#   afk-iter-packages-plugin-commits-without-changesets-orchestrator-main-turn-back-fill-is-fragile-recovery-hook-level-enforcement-preferable-problem     — sibling changeset-discipline hook (candidate).
+#   readme-refresh-enforcement-gap-iter-subprocess-commits-can-land-a-verifying-md-rename-without-staging-the-corresponding-verification-queue-row-in-docs-problems-readme-md-problem     — sibling README-refresh-discipline hook (first consumer
+#              under hook-substring-matches-git-commit-anywhere-in-bash-command-not-just-actual-git-commit-invocations-problem).
+#   hook-substring-matches-git-commit-anywhere-in-bash-command-not-just-actual-git-commit-invocations-problem     — this helper's surfacing ticket.
 
 # Returns 0 if $1 is a Bash command that invokes `git commit`.
 # Returns 1 otherwise (including commands that merely mention the

package/hooks/lib/create-gate.sh

@@ -1,5 +1,5 @@
 #!/bin/bash
-# Shared gate logic for new-ticket creation enforcement (P119).
+# Shared gate logic for new-ticket creation enforcement (agent-bypasses-wr-itil-manage-problem-step-2-duplicate-check-by-writing-tickets-directly-to-docs-problems-problem).
 #
 # Sourced by manage-problem-enforce-create.sh. Provides:
 #   check_create_gate     — returns 0 if Step 2 marker present (allow), 1 if absent (deny)
@@ -12,7 +12,7 @@
 # marker; mismatch invalidates the gate). The Step-2 grep marker has
 # different semantics — it records "this session has run the duplicate
 # check at least once" — and has neither a policy file nor drift-relevant
-# state. Per architect approval (P119): keep review-gate.sh untouched and
+# state. Per architect approval (agent-bypasses-wr-itil-manage-problem-step-2-duplicate-check-by-writing-tickets-directly-to-docs-problems-problem): keep review-gate.sh untouched and
 # add a sibling helper rather than overload review-gate semantics.
 #
 # Marker convention: /tmp/manage-problem-grep-${SESSION_ID}
@@ -23,9 +23,9 @@
 # scope would block split-create after the first Write.
 #
 # References:
-#   ADR-009 — gate marker lifecycle (covers session-scoped /tmp markers).
-#   ADR-038 — progressive disclosure (deny message stays terse + actionable).
-#   ADR-013 Rule 1 — deny redirects to /wr-itil:manage-problem where
+#   gate-marker-lifecycle-ttl-drift-not-stop-hook-reset-architecture-rule — gate marker lifecycle (covers session-scoped /tmp markers).
+#   progressive-disclosure-once-per-session-budget-for-userpromptsubmit-governance-prose-architecture-rule — progressive disclosure (deny message stays terse + actionable).
+#   structured-user-interaction-for-governance-skill-decisions-architecture-rule Rule 1 — deny redirects to /wr-itil:manage-problem where
 #                    Step 2 fires AskUserQuestion if duplicates exist.
 #
 # Empty SESSION_ID fallback: returns 1 (no marker) — fail-closed by
@@ -52,19 +52,19 @@ mark_step2_complete() {
   : > "/tmp/manage-problem-grep-${SESSION_ID}"
 }
 
-# P260 / ADR-050 Option C: write the Step 2 grep marker under EVERY
+# create-gate-marker-race-between-concurrent-claude-sessions-via-shared-runtime-sid-file-problem / capture-the-runtime-stdin-session-id-via-a-pretooluse-hook-so-the-create-gate-marker-binds-to-the-same-sid-the-runtime-hook-will-see-architecture-rule Option C: write the Step 2 grep marker under EVERY
 # candidate SID read from stdin (one UUID per line, as emitted by
 # `get_candidate_session_ids` in lib/session-id.sh).
 #
 # Under concurrent orchestrator+subprocess sessions in the same project,
 # the per-machine runtime-sid marker is last-writer-wins, so agent-side
 # code cannot reliably PREDICT which single SID the create-gate hook will
-# read from the Write's stdin (ADR-050 §Context). Marking under every recent
+# read from the Write's stdin (capture-the-runtime-stdin-session-id-via-a-pretooluse-hook-so-the-create-gate-marker-binds-to-the-same-sid-the-runtime-hook-will-see-architecture-rule §Context). Marking under every recent
 # candidate guarantees a matching marker exists whichever SID the hook reads,
-# eliminating the P260 marker-mismatch deny without a process-topology
+# eliminating the create-gate-marker-race-between-concurrent-claude-sessions-via-shared-runtime-sid-file-problem marker-mismatch deny without a process-topology
 # assumption. The candidate set is bounded to recent announce markers + the
 # runtime-sid value by `get_candidate_session_ids` (NOT a global fail-open —
-# the P119 audit invariant holds: each marker still records that THIS session
+# the agent-bypasses-wr-itil-manage-problem-step-2-duplicate-check-by-writing-tickets-directly-to-docs-problems-problem audit invariant holds: each marker still records that THIS session
 # ran the duplicate-check grep).
 #
 # Reuses the unchanged single-SID `mark_step2_complete` per candidate (same
@@ -85,36 +85,6 @@ mark_step2_complete_candidates() {
   [ "$count" -gt 0 ]
 }
 
-# Returns 0 if the RFC capture-step marker exists for SESSION_ID; 1 otherwise.
-# Empty SESSION_ID => returns 1 (no marker).
-#
-# Sibling marker per architect verdict on capture-rfc sub-decision (a) —
-# preserves audit-trail per-surface granularity (problem-tier vs RFC-tier
-# capture). Marker name: /tmp/wr-itil-rfc-capture-grep-${SESSION_ID}.
-#
-# Usage: if check_rfc_capture_gate "$SESSION_ID"; then exit 0; fi
-check_rfc_capture_gate() {
-  local SESSION_ID="$1"
-  [ -n "$SESSION_ID" ] || return 1
-  [ -f "/tmp/wr-itil-rfc-capture-grep-${SESSION_ID}" ]
-}
-
-# Writes the RFC capture-step marker for SESSION_ID. Empty SESSION_ID => no-op.
-# Idempotent — safe to call more than once per session.
-#
-# Per ADR-060 + capture-rfc Step 2: the marker records that capture-rfc has
-# run its problem-trace validation pass (the RFC-tier analogue of the
-# manage-problem Step 2 duplicate-grep). Per-session scope so a single
-# capture-rfc invocation may write multiple RFC files (e.g. multi-problem
-# trace splits) without re-validating.
-#
-# Usage: mark_rfc_capture_complete "$SESSION_ID"
-mark_rfc_capture_complete() {
-  local SESSION_ID="$1"
-  [ -n "$SESSION_ID" ] || return 0
-  : > "/tmp/wr-itil-rfc-capture-grep-${SESSION_ID}"
-}
-
 # Emit fail-closed deny JSON for PreToolUse hooks.
 # Usage: create_gate_deny "BLOCKED: <reason>"
 create_gate_deny() {

package/hooks/lib/detectors.sh

@@ -1,18 +1,18 @@
 #!/bin/bash
-# P085 detector registry: assistant-output-gate + assistant-output-review.
+# problem-085-assistant-asks-for-input-when-the-next-step-is-obvious-and-uses-prose-asks-instead-of-askuserquestion-when-input-is-actually-needed-problem detector registry: assistant-output-gate + assistant-output-review.
 #
 # Two detection functions sharing a single library so the UserPromptSubmit
 # pre-generation reminder (gate) and the Stop post-hoc review use the
 # same canonical phrasing list — one place to update when a new prose-ask
 # pattern lands.
 #
-# Composition: this registry is the shape P078 (correction->ticket) and
+# Composition: this registry is the shape problem-078-assistant-does-not-offer-to-capture-a-problem-ticket-when-the-user-delivers-strong-signal-correction-problem (correction->ticket) and
 # future itil assistant-output validators extend. Each detector is a
 # pure function — takes text on stdin or as $1, exits 0 on match,
 # non-zero on no-match.
 
 # Canonical prose-ask phrasings — the patterns that should be emitted
-# via AskUserQuestion instead. Extracted from P085 ticket + memory
+# via AskUserQuestion instead. Extracted from problem-085-assistant-asks-for-input-when-the-next-step-is-obvious-and-uses-prose-asks-instead-of-askuserquestion-when-input-is-actually-needed-problem ticket + memory
 # feedback_act_on_obvious_decisions.md.
 #
 # Case-insensitive, anchored to word boundaries where meaningful.
@@ -57,12 +57,12 @@ DIRECTION_PIN_PATTERNS=(
 
 # Correction-signal patterns — strong-affect correction signals in
 # the user's incoming prompt that should trigger an offer to capture
-# a problem ticket for the underlying behavioural pattern (P078).
-# Extracted from feedback_capture_on_correction.md and the P078 ticket
+# a problem ticket for the underlying behavioural pattern (problem-078-assistant-does-not-offer-to-capture-a-problem-ticket-when-the-user-delivers-strong-signal-correction-problem).
+# Extracted from feedback_capture_on_correction.md and the problem-078-assistant-does-not-offer-to-capture-a-problem-ticket-when-the-user-delivers-strong-signal-correction-problem ticket
 # Root Cause Analysis "missing trigger" vocabulary.
 #
 # Case-insensitive (grep -Eqi). False-positive budget is accepted per
-# P078 Investigation Tasks (a user message quoting the vocabulary in
+# problem-078-assistant-does-not-offer-to-capture-a-problem-ticket-when-the-user-delivers-strong-signal-correction-problem Investigation Tasks (a user message quoting the vocabulary in
 # non-correction context will fire — minor over-report traded for
 # signal reliability).
 CORRECTION_SIGNAL_PATTERNS=(
@@ -79,22 +79,22 @@ CORRECTION_SIGNAL_PATTERNS=(
   '\bno\b.*\bwrong\b'
 )
 
-# P403: mechanical-step-framed-as-user-optional detection. When a skill
+# agent-frames-skill-mandatory-mechanical-steps-as-user-optional-problem: mechanical-step-framed-as-user-optional detection. When a skill
 # contract mandates a mechanical step (review-problems Step 7 auto-release,
 # a full re-rank, etc.), the agent must not re-surface the step as a user
 # decision — or as a budget-caution skip — in end-of-turn prose. That
-# reintroduces the exact friction the mechanical-stage carve-out (P132 /
-# ADR-044 category 4 silent-framework) was engineered to remove.
+# reintroduces the exact friction the mechanical-stage carve-out (agents-over-ask-in-interactive-sessions-conflating-mechanical-stages-with-user-interactive-stages-of-multi-stage-skill-contracts-inverse-problem /
+# decision-delegation-contract-when-agents-act-on-the-framework-vs-ask-the-user-architecture-rule category 4 silent-framework) was engineered to remove.
 #
 # The detector fires only when ALL THREE co-occur in the turn text: a
 # step-skip verb, a step / mechanical-pass reference, AND an
 # offloaded-justification closer. The three-way AND is the anti-over-fire
-# guard (inverse-P078 / P132): a legitimately user-owned decision, or a
+# guard (inverse-problem-078-assistant-does-not-offer-to-capture-a-problem-ticket-when-the-user-delivers-strong-signal-correction-problem / agents-over-ask-in-interactive-sessions-conflating-mechanical-stages-with-user-interactive-stages-of-multi-stage-skill-contracts-inverse-problem): a legitimately user-owned decision, or a
 # skipped step reported with a substantive reason, does not fire. Same
 # accepted false-positive posture as CORRECTION_SIGNAL_PATTERNS above — a
 # rare over-fire costs only a non-blocking next-turn nudge.
 #
-# Case-insensitive (grep -Eqi). grep matches per line; the two P403
+# Case-insensitive (grep -Eqi). grep matches per line; the two agent-frames-skill-mandatory-mechanical-steps-as-user-optional-problem
 # evidence phrasings are single-line, so per-line matching is sufficient.
 
 # 1. Step-skip verbs — the step was not run.
@@ -196,7 +196,7 @@ detect_direction_pin() {
 }
 
 # detect_correction_signal: scans text on stdin for strong-affect
-# correction signals (P078). Exits 0 if any pattern matches, 1
+# correction signals (problem-078-assistant-does-not-offer-to-capture-a-problem-ticket-when-the-user-delivers-strong-signal-correction-problem). Exits 0 if any pattern matches, 1
 # otherwise. Writes the first matched phrase to stdout for
 # observability in the hook's systemMessage.
 #