@windyroad/itil 0.61.1 → 0.61.2

package/skills/work-problems/test/work-problems-step-5-stream-timeout-salvage.bats

@@ -1,251 +0,0 @@
-#!/usr/bin/env bats
-# tdd-review: structural-permitted (justification: the doc-lint slice below
-# asserts SKILL.md / ADR-032 prose contract — SKILL.md is the contract
-# document per ADR-037 Permitted Exception; these guards catch prose drift
-# away from the behavioural SALVAGE/HALT contract exercised above. The
-# load-bearing core of this fixture is behavioural per ADR-052. harness-gap P012)
-#
-# Behavioural test: work-problems Step 5 exit-code semantics — the is_error:true
-# stream-timeout SALVAGE carve-out (P261). When an iter subprocess returns
-# `is_error: true` (e.g. `API Error: Stream idle timeout - partial response
-# received`) AFTER staging coherent work but BEFORE `git commit`, the staged
-# work survives in the working tree. The orchestrator's salvage decision logic
-# is: IF is_error:true AND staged files exist AND iter-authored bats pass →
-# SALVAGE (commit the staged work from the main turn with iter-attribution; the
-# commit gate fires fresh). ELSE → HALT per the existing exit-code contract.
-#
-# This is a NEW recovery branch, distinct from:
-#   - P121 (SIGTERM idle-timeout — is_error:false clean exit-flush; subprocess
-#     HAD committed before going idle)
-#   - P147 (SIGTERM stuck-before-emit — exit 143 + 0-byte JSON; metadata lost)
-#   - P146 (bash-polling antipattern — the deadlock mechanism behind P147)
-# The stream-timeout class preserves metadata in the JSON envelope AND the
-# staged files; the iter exits on its own with is_error:true (no SIGTERM).
-#
-# The fake-stuck-shim below re-creates the production shape: it stages coherent
-# work in the repo, then emits an is_error:true stream-timeout JSON envelope
-# (no ITERATION_SUMMARY, no commit). The harness re-implements the orchestrator's
-# SALVAGE-vs-HALT decision contract (faithful to SKILL.md Step 5) and asserts the
-# branch outcome across the four input shapes. Adopters who copy the SKILL.md
-# Step 5 exit-code block into their orchestrator should observe the same outcomes.
-#
-# @problem P261
-# @jtbd JTBD-006
-# @jtbd JTBD-001
-#
-# Cross-reference:
-#   P261 (iter subprocess API stream-timeout salvage path) — driver ticket
-#   ADR-032 (governance skill invocation patterns — is_error:true stream-timeout
-#     salvage sub-variant, P261 amendment) — the carved-out commit-authorship
-#     contract this fixture pins
-#   ADR-009 (gate-marker-lifecycle — is_error:true subprocess MUST NOT extend
-#     parent trust window; salvage commit fires the gate fresh on the
-#     orchestrator's own SESSION_ID)
-#   ADR-014 (single-commit grain — the salvage commit IS the iteration's one
-#     commit; amend-folding is inapplicable because no iter commit exists)
-#   ADR-037 / ADR-052 (skill testing strategy — behavioural default; doc-lint
-#     contract assertion is the Permitted Exception, marked above)
-
-setup() {
-  TEST_TMP="$(mktemp -d)"
-  FAKE_BIN="${TEST_TMP}/bin"
-  mkdir -p "$FAKE_BIN"
-
-  # Fake `claude` binary simulating a stuck iteration subprocess of the
-  # stream-timeout class: it stages coherent work in the CWD git repo, then
-  # emits an is_error:true JSON envelope carrying the stream-timeout error
-  # string in `.result`. No ITERATION_SUMMARY; no commit. This matches the
-  # 2026-05-18 session-6 iter-4 shape captured in P261: 7 files staged, then
-  # `API Error: Stream idle timeout - partial response received`.
-  cat > "$FAKE_BIN/claude" <<'FAKE_EOF'
-#!/usr/bin/env bash
-# Test fake for work-problems Step 5 P261 stream-timeout salvage fixture.
-# Stages coherent work, then emits is_error:true stream-timeout JSON.
-if [ "${FAKE_STAGE_WORK:-1}" = "1" ]; then
-  printf 'salvaged SKILL amendment + bats fixture\n' > staged-iter-work.txt
-  git add staged-iter-work.txt 2>/dev/null || true
-fi
-if [ "${FAKE_IS_ERROR:-true}" = "true" ]; then
-  printf '%s\n' '{"is_error":true,"result":"API Error: Stream idle timeout - partial response received","total_cost_usd":12.91,"duration_ms":300000,"usage":{"input_tokens":1000,"output_tokens":2000,"cache_creation_input_tokens":0,"cache_read_input_tokens":0}}'
-else
-  printf '%s\n' '{"is_error":false,"result":"ITERATION_SUMMARY\nticket_id: P000\naction: worked\noutcome: investigated\ncommitted: true\nremaining_backlog_count: 0\nnotes: normal exit","total_cost_usd":0.5,"duration_ms":1000,"usage":{"input_tokens":10,"output_tokens":20,"cache_creation_input_tokens":0,"cache_read_input_tokens":0}}'
-fi
-FAKE_EOF
-  chmod +x "$FAKE_BIN/claude"
-
-  # Fake `iter_bats` stub — stands in for running the iter-authored bats
-  # fixtures as the salvage path's step-1 structural sanity check. Its exit
-  # code is controlled by FAKE_BATS_EXIT (0 = green, 1 = fail) so the harness
-  # can exercise both the bats-green and bats-fail branches behaviourally.
-  cat > "$FAKE_BIN/iter_bats" <<'FAKE_EOF'
-#!/usr/bin/env bash
-exit "${FAKE_BATS_EXIT:-0}"
-FAKE_EOF
-  chmod +x "$FAKE_BIN/iter_bats"
-  export PATH="$FAKE_BIN:$PATH"
-
-  # A throwaway git repo so staged-work detection + the salvage commit are real.
-  REPO="${TEST_TMP}/repo"
-  mkdir -p "$REPO"
-  git -C "$REPO" init -q
-  git -C "$REPO" config user.email "test@example.com"
-  git -C "$REPO" config user.name "Test"
-  git -C "$REPO" commit -q --allow-empty -m "root"
-
-  SKILL_DIR="$(cd "$(dirname "$BATS_TEST_FILENAME")/.." && pwd)"
-  SKILL_FILE="${SKILL_DIR}/SKILL.md"
-  ADR_FILE="$(cd "${SKILL_DIR}/../../../.." && pwd)/docs/decisions/032-governance-skill-invocation-patterns.proposed.md"
-}
-
-teardown() {
-  if [ -n "${TEST_TMP:-}" ] && [ -d "$TEST_TMP" ]; then
-    rm -rf "$TEST_TMP"
-  fi
-}
-
-# Faithful re-implementation of SKILL.md Step 5's is_error:true salvage decision
-# contract. Consumes the iter JSON envelope + the repo working-tree state and
-# returns the orchestrator's branch decision. The SALVAGE branch performs the
-# real 4-step path's commit (step 3) so the commit is observable; the bats
-# sanity check (step 1) is exercised via the FAKE_BATS_EXIT-controlled stub.
-salvage_decision() {
-  local json="$1"
-  local repo="$2"
-
-  local is_error
-  is_error=$(printf '%s' "$json" | python3 -c 'import json,sys; print(str(json.load(sys.stdin).get("is_error")).lower())')
-
-  # is_error:false → normal exit-code path; not the salvage branch.
-  if [ "$is_error" != "true" ]; then
-    printf 'DECISION=PARSE_SUMMARY\n'
-    return 0
-  fi
-
-  # is_error:true with no staged work → halt per the existing exit-code contract.
-  local staged
-  staged=$(git -C "$repo" diff --cached --name-only)
-  if [ -z "$staged" ]; then
-    printf 'DECISION=HALT reason=no-staged-work\n'
-    return 0
-  fi
-
-  # is_error:true with staged work → run the iter-authored bats as the
-  # structural sanity check (step 1). Green → SALVAGE; fail → HALT.
-  if iter_bats; then
-    # Step 3: commit the staged work from the orchestrator main turn with
-    # explicit iter-attribution. (Step 4 — fresh commit gate — is the runtime
-    # orchestrator's concern, asserted via the doc-lint slice below.)
-    git -C "$repo" commit -q -m "salvage(P000): iter hit API stream timeout before commit — committed staged work from orchestrator main turn; iter-authored bats green"
-    printf 'DECISION=SALVAGE\n'
-  else
-    printf 'DECISION=HALT reason=bats-fail\n'
-  fi
-  return 0
-}
-
-# ---------------------------------------------------------------------------
-# Behavioural cases (the load-bearing core per ADR-052).
-# ---------------------------------------------------------------------------
-
-@test "P261: is_error:true + staged work + bats green -> SALVAGE (commit from main turn)" {
-  export FAKE_IS_ERROR=true FAKE_STAGE_WORK=1 FAKE_BATS_EXIT=0
-  local json
-  json=$( cd "$REPO" && claude -p --output-format json "TEST" < /dev/null )
-  run salvage_decision "$json" "$REPO"
-  [ "$status" -eq 0 ]
-  [[ "$output" == *"DECISION=SALVAGE"* ]]
-  # The salvage commit must have landed, carrying the staged work + attribution.
-  run git -C "$REPO" log -1 --format=%s
-  [[ "$output" == *"salvage"* ]]
-  [[ "$output" == *"orchestrator main turn"* ]]
-  run git -C "$REPO" show --stat HEAD
-  [[ "$output" == *"staged-iter-work.txt"* ]]
-}
-
-@test "P261: is_error:true + staged work + bats FAIL -> HALT (incoherent work not salvaged)" {
-  export FAKE_IS_ERROR=true FAKE_STAGE_WORK=1 FAKE_BATS_EXIT=1
-  local json
-  json=$( cd "$REPO" && claude -p --output-format json "TEST" < /dev/null )
-  run salvage_decision "$json" "$REPO"
-  [ "$status" -eq 0 ]
-  [[ "$output" == *"DECISION=HALT"* ]]
-  [[ "$output" == *"reason=bats-fail"* ]]
-  # No salvage commit landed — HEAD is still the root commit.
-  run git -C "$REPO" log --oneline
-  [ "$(printf '%s\n' "$output" | grep -c .)" -eq 1 ]
-}
-
-@test "P261: is_error:true + NO staged work -> HALT per existing exit-code contract" {
-  export FAKE_IS_ERROR=true FAKE_STAGE_WORK=0 FAKE_BATS_EXIT=0
-  local json
-  json=$( cd "$REPO" && claude -p --output-format json "TEST" < /dev/null )
-  run salvage_decision "$json" "$REPO"
-  [ "$status" -eq 0 ]
-  [[ "$output" == *"DECISION=HALT"* ]]
-  [[ "$output" == *"reason=no-staged-work"* ]]
-}
-
-@test "P261: is_error:false (normal exit) -> PARSE_SUMMARY, NOT the salvage branch" {
-  export FAKE_IS_ERROR=false FAKE_STAGE_WORK=0
-  local json
-  json=$( cd "$REPO" && claude -p --output-format json "TEST" < /dev/null )
-  run salvage_decision "$json" "$REPO"
-  [ "$status" -eq 0 ]
-  [[ "$output" == *"DECISION=PARSE_SUMMARY"* ]]
-}
-
-# ---------------------------------------------------------------------------
-# Doc-lint contract assertions (Permitted Exception per ADR-037; structural
-# slice marked at top of file per ADR-052 Surface 2). These guard the SKILL.md
-# / ADR-032 prose against drift away from the behavioural contract above.
-# ---------------------------------------------------------------------------
-
-@test "P261: SKILL.md Step 5 exit-code semantics documents the is_error:true salvage carve-out" {
-  run grep -niE "is_error.{0,30}salvage|salvage.{0,40}is_error|stream.?idle.?timeout.{0,80}salvage|salvage.{0,80}stream.?(idle.?)?timeout" "$SKILL_FILE"
-  [ "$status" -eq 0 ]
-}
-
-@test "P261: SKILL.md salvage carve-out names the staged-work + bats-pass gate condition" {
-  run grep -niE "staged.{0,40}(file|work).{0,80}bats|bats.{0,80}(pass|green).{0,80}salvage|salvage.{0,120}staged" "$SKILL_FILE"
-  [ "$status" -eq 0 ]
-}
-
-@test "P261: SKILL.md salvage carve-out documents the commit-from-main-turn + fresh-gate steps" {
-  run grep -niE "(orchestrator|main turn).{0,80}commit.{0,120}(attribut|fresh)|commit gate fires fresh|fresh.{0,30}commit gate" "$SKILL_FILE"
-  [ "$status" -eq 0 ]
-}
-
-@test "P261: SKILL.md salvage carve-out distinguishes the class from P147 / P121 / P146" {
-  run grep -niE "P147" "$SKILL_FILE"
-  [ "$status" -eq 0 ]
-  run grep -niE "P146" "$SKILL_FILE"
-  [ "$status" -eq 0 ]
-}
-
-@test "P261: SKILL.md Step 5 cites P261 (salvage-carve-out driver)" {
-  run grep -nE "P261" "$SKILL_FILE"
-  [ "$status" -eq 0 ]
-}
-
-@test "P261: SKILL.md line ~486 orchestrator-commit rule carries the salvage exception cross-reference" {
-  # The 'orchestrator does NOT commit from its main turn' rule must no longer
-  # read as unqualified — it must name the salvage carve-out as the one
-  # bounded exception so SKILL.md and ADR-032 stay in agreement.
-  run grep -niE "orchestrator does NOT commit from its main turn.{0,200}(except|salvage)|salvage.{0,80}(except|one case).{0,120}main turn" "$SKILL_FILE"
-  [ "$status" -eq 0 ]
-}
-
-@test "P261: ADR-032 carries the is_error:true stream-timeout salvage sub-variant amendment" {
-  run grep -niE "is_error.{0,30}(stream.?timeout)?.{0,30}salvage|stream.?timeout salvage|P261 amendment" "$ADR_FILE"
-  [ "$status" -eq 0 ]
-}
-
-@test "P261: ADR-032 salvage amendment preserves one-commit-per-iteration grain (amend-folding inapplicable)" {
-  run grep -niE "amend.?(based)?.?folding.{0,80}(inapplicable|no iter commit|not apply)|salvage commit IS the iteration|no iter commit.{0,40}amend" "$ADR_FILE"
-  [ "$status" -eq 0 ]
-}
-
-@test "P261: ADR-032 salvage amendment confirms fresh-gate-marker behaviour per ADR-009" {
-  run grep -niE "ADR-009|fresh.{0,30}(gate|commit gate)|own SESSION_ID|trust window" "$ADR_FILE"
-  [ "$status" -eq 0 ]
-}

package/skills/work-problems/test/work-problems-step-6-5-always-drain.bats

@@ -1,203 +0,0 @@
-#!/usr/bin/env bats
-
-# P250: /wr-itil:work-problems Step 6.5 release-cadence classification must
-# pivot on releasable material, not residual band. The defective prior
-# clause "Within appetite (≤ 3/25) — no drain needed" encoded
-# accumulation-permitted-below-threshold semantics that violated the
-# user's verbatim direction: "If it's low risk, you should release."
-#
-# Amended contract (three-band):
-#   1. Above appetite (≥ 5/25)        → ADR-042 auto-apply (unchanged).
-#   2. Within appetite (≤ 4/25) AND releasable material → drain.
-#   3. Within appetite (≤ 4/25) AND empty queue        → no drain (no-op fast-path).
-#
-# The trigger for the drain action is *presence of releasable material*
-# (any unpushed commits OR any .changeset/ entries). The residual band remains
-# the safety check (above-appetite never releases) but is no longer the
-# action gate for the within-appetite branch.
-#
-# Doc-lint contract assertions per ADR-037 Permitted Exception (contract-
-# assertion class). The asserted prose IS the load-bearing policy surface
-# — re-reading SKILL.md is the only way an AFK reader (and the iteration
-# subprocess) learns the new classification. Behavioural verification
-# would require executing the orchestrator's decision tree, which is not
-# scriptable until the Phase 2 advisory-classifier from P081 ships. These
-# tests function as regression guards against re-introducing the
-# "Within appetite (≤ 3/25) — no drain needed" wording or any equivalent
-# threshold-as-action-gate framing.
-#
-# @problem P250
-# @adr ADR-018 (release-cadence policy parent — amended in same commit)
-# @adr ADR-037 (skill-testing strategy — contract-assertion class)
-# @adr ADR-042 (above-appetite branch — preserved unchanged)
-# @jtbd JTBD-006 (Progress the Backlog While I'm Away — primary)
-# @jtbd JTBD-002 (Ship with Confidence — composes; small frequent releases)
-
-setup() {
-  REPO_ROOT="$(cd "$(dirname "$BATS_TEST_FILENAME")/../../../../.." && pwd)"
-  SKILL_MD="$REPO_ROOT/packages/itil/skills/work-problems/SKILL.md"
-  ADR_018="$REPO_ROOT/docs/decisions/018-inter-iteration-release-cadence-for-afk-loops.proposed.md"
-}
-
-# ── Preconditions ──────────────────────────────────────────────────────────
-
-@test "work-problems P250: SKILL.md exists" {
-  [ -f "$SKILL_MD" ]
-}
-
-@test "work-problems P250: ADR-018 exists" {
-  [ -f "$ADR_018" ]
-}
-
-# ── Regression guard: defective prior wording is gone ──────────────────────
-
-@test "work-problems P250: SKILL.md no longer contains 'Within appetite (≤ 3/25) — no drain needed' clause" {
-  # The defective clause. Re-introducing it regresses to accumulation-
-  # permitted-below-threshold semantics.
-  run grep -nE 'Within appetite \(≤ 3/25\).*no drain needed' "$SKILL_MD"
-  [ "$status" -ne 0 ]
-}
-
-@test "work-problems P250: SKILL.md no longer treats '= 4/25' as a discrete drain trigger band" {
-  # The prior "At appetite (= 4/25) — drain" band is collapsed into the
-  # new "≤ 4/25 AND releasable material" branch. The discrete = 4/25
-  # framing should be absent from the Step 6.5 classification.
-  run grep -nE 'At appetite \(= 4/25\) — drain' "$SKILL_MD"
-  [ "$status" -ne 0 ]
-}
-
-# ── Above-appetite branch preserved (ADR-042) ──────────────────────────────
-
-@test "work-problems P250: Above-appetite (≥ 5/25) branch routes to ADR-042 auto-apply (preserved)" {
-  run grep -nE 'Above appetite \(≥ 5/25\).*Above-appetite branch' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-@test "work-problems P250: Above-appetite branch references the auto-apply loop (ADR-042 intact)" {
-  run grep -nE 'auto-apply loop|ADR-042' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-# ── Within-appetite + releasable material drains ──────────────────────────
-
-@test "work-problems P250: Within-appetite + releasable material triggers drain" {
-  # The load-bearing positive contract: within appetite AND any unpushed
-  # commits OR changeset entry → drain.
-  run grep -nE 'Within appetite \(≤ 4/25\) AND there is releasable material' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-@test "work-problems P250: releasable-material clause enumerates unpushed commits" {
-  run grep -nE 'releasable material.*unpushed commits|unpushed commits.*releasable material|any unpushed commits' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-@test "work-problems P250: releasable-material clause enumerates .changeset/ entries" {
-  run grep -nE 'any entries in `\.changeset/`|`\.changeset/` non-empty|entries in `\.changeset/`' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-# ── Within-appetite + empty queue no-op fast-path ─────────────────────────
-
-@test "work-problems P250: Within-appetite + empty queue does NOT drain (no-op fast-path)" {
-  # The genuine fast-path: nothing to release. Gate is *absence of
-  # releasable material*, not residual band.
-  run grep -nE 'Within appetite \(≤ 4/25\) AND empty queue' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-@test "work-problems P250: empty-queue branch is framed as 'literally nothing to release', not a threshold deferral" {
-  # Regression guard: the no-drain branch must NOT be reframed as a
-  # threshold-based defer. The gate is queue emptiness, not residual.
-  run grep -nE 'literally nothing to release|nothing to release.*genuine no-op' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-# ── User-direction citation preserved ─────────────────────────────────────
-
-@test "work-problems P250: amended classification cites user direction verbatim ('If it's low risk, you should release.')" {
-  # The user's verbatim direction is the load-bearing rationale. Future
-  # readers tracing the amendment back must find the citation without
-  # keyword-guessing.
-  run grep -nE 'If it.s low risk, you should release' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-@test "work-problems P250: classification cites P250 by ticket ID" {
-  # ADR-022 amendment trail: the amended clause must self-identify
-  # so future readers tracing back from the ticket find it.
-  run grep -nE 'P250' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-# ── Non-Interactive Decision Making table reflects amendment ──────────────
-
-@test "work-problems P250: Decision Making table carries the within-appetite-with-releasable-material row" {
-  # The decision table is the AFK reader's quick-reference; the
-  # amendment must surface here too, otherwise the table contradicts
-  # Step 6.5's classification.
-  run grep -nE '\| Pipeline risk within appetite \(≤ 4/25\) with releasable material' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-@test "work-problems P250: Decision Making table carries the empty-queue no-drain row" {
-  run grep -nE '\| Pipeline risk within appetite \(≤ 4/25\) AND empty queue' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-@test "work-problems P250: Decision Making table row cites 'presence of releasable material', not 'residual reaching appetite'" {
-  # The trigger semantics must be explicit in the table row, not
-  # buried in Step 6.5 prose 100+ lines above.
-  run grep -nE 'presence of releasable material|releasable material.*not residual band|Trigger is.*presence' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-# ── ADR-018 amendment (in same commit per ADR-014) ────────────────────────
-
-@test "work-problems P250: ADR-018 contains the 2026-05-18 Drain-trigger amendment heading" {
-  # ADR-014 single-unit-of-work: the ADR amendment lands in the same
-  # commit as the SKILL change. ADR-018's Mechanism without this
-  # amendment contradicts SKILL.md Step 6.5.
-  run grep -nE 'Amendment 2026-05-18.*Drain trigger is releasable material' "$ADR_018"
-  [ "$status" -eq 0 ]
-}
-
-@test "work-problems P250: ADR-018 amendment cites P250 by ticket ID" {
-  run grep -nE 'P250' "$ADR_018"
-  [ "$status" -eq 0 ]
-}
-
-@test "work-problems P250: ADR-018 amendment quotes the user direction" {
-  run grep -nE 'If it.s low risk, you should release' "$ADR_018"
-  [ "$status" -eq 0 ]
-}
-
-@test "work-problems P250: ADR-018 amendment encodes the new drain condition shape" {
-  # The amendment must contain the conjunction "≤ 4/25 AND (releasable
-  # material disjunct)" so future readers tracing the policy chain
-  # find the rule in the load-bearing ADR, not just in the SKILL.
-  run grep -nE 'residual ≤ 4/25 AND' "$ADR_018"
-  [ "$status" -eq 0 ]
-}
-
-@test "work-problems P250: ADR-018 amendment preserves Above-appetite delegation to ADR-042" {
-  # The amendment MUST NOT touch the above-appetite invariant. ADR-042
-  # remains the safety gate.
-  run grep -nE 'Above-appetite states \(≥ 5/25\) route to ADR-042|Above-appetite.*ADR-042' "$ADR_018"
-  [ "$status" -eq 0 ]
-}
-
-# ── Cross-references intact ───────────────────────────────────────────────
-
-@test "work-problems P250: SKILL.md Step 6.5 still cites ADR-018 (parent policy intact)" {
-  run grep -nE 'Step 6\.5.*ADR-018|ADR-018.*Step 6\.5|Release-cadence check.*ADR-018' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-@test "work-problems P250: SKILL.md Step 6.5 still cites P041 (accumulation root cause cross-reference)" {
-  # P041 is the ancestor that established Step 6.5 itself. The
-  # amendment refines P041's solution; the cross-reference must
-  # survive.
-  run grep -nE 'P041' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}

package/skills/work-problems/test/work-problems-step-6-5-cache-refresh-chain.bats

@@ -1,174 +0,0 @@
-#!/usr/bin/env bats
-
-# P233: /wr-itil:work-problems Step 6.5 Drain action must chain
-# /install-updates AFTER successful release:watch so the next iter
-# subprocess loads the just-shipped plugin from cache rather than the
-# pre-release cached version.
-#
-# Empirical driver (briefing/afk-subprocess.md:18): iter subprocesses
-# re-resolve plugin cache on spawn; without cache refresh between
-# release:watch and next-iter dispatch, the just-shipped hook is
-# inactive in the next iter — defeating the "ship a hook to prevent
-# recurrence" pattern for the immediate-next-iter case.
-#
-# Doc-lint contract assertions per ADR-037 Permitted Exception
-# (contract-assertion class — same shape as the P140 / P130 / P126 / P135
-# sibling fixtures). The asserted prose IS the load-bearing policy
-# surface — the chained invocation is documentation-only (the orchestrator
-# is a prose-driven agent, not a script); the SKILL.md is the contract.
-# Behavioural assertion would require a fixture project + real claude -p
-# subprocess that observes a freshly-installed plugin — outside scope.
-#
-# @problem P233
-# @adr ADR-013 (Rule 5 — policy-authorised silent action; chains the
-#       same authorisation that covers push:watch / release:watch)
-# @adr ADR-014 (one-commit-per-iter; this edit lands in one commit)
-# @adr ADR-018 (inter-iteration release cadence; this refines its
-#       Drain action sub-block with the post-release cache refresh)
-# @adr ADR-030 (repo-local skills — install-updates is the chained
-#       skill)
-# @adr ADR-037 (skill-testing strategy — contract-assertion class)
-# @adr ADR-044 (decision-delegation contract — framework-resolution
-#       boundary; install-updates' AskUserQuestion-bearing branches
-#       fall through to its Non-interactive fallback under P130)
-# @jtbd JTBD-006 (Progress the Backlog While I'm Away — primary;
-#       outcome 7 "risk never silently accumulates across AFK iterations"
-#       extends to "just-shipped hooks effective on next iter")
-# @jtbd JTBD-001 (Enforce Governance Without Slowing Down — composes;
-#       just-shipped governance hooks become effective immediately,
-#       not after manual /install-updates re-run)
-# @jtbd JTBD-007 (Keep Plugins Current Across Projects — persona-level
-#       currency anchor that JTBD-006's AFK use case extends)
-
-setup() {
-  REPO_ROOT="$(cd "$(dirname "$BATS_TEST_FILENAME")/../../../../.." && pwd)"
-  SKILL_MD="$REPO_ROOT/packages/itil/skills/work-problems/SKILL.md"
-}
-
-@test "work-problems P233: SKILL.md exists" {
-  [ -f "$SKILL_MD" ]
-}
-
-# ── Cache-refresh chain identity ───────────────────────────────────────────
-
-@test "work-problems P233: Step 6.5 Drain action carries a post-release cache-refresh subsection citing P233" {
-  # The amendment must self-identify so future readers tracing back from
-  # the ticket find the load-bearing prose without keyword-guessing.
-  run grep -nE 'cache.refresh.*P233|P233.*cache.refresh|post.release.cache.*P233|P233.*post.release.cache' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-@test "work-problems P233: Drain action chains /install-updates after successful release:watch" {
-  # Core contract: after release:watch returns success, the orchestrator
-  # invokes /install-updates to refresh the plugin cache. Without this
-  # chain, the next iter subprocess loads the pre-release cached version
-  # (P233's empirical root cause).
-  run grep -nE '/install-updates|install-updates skill' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-@test "work-problems P233: cache-refresh is conditional on actual release (skipped when no changeset)" {
-  # The chain MUST NOT fire when release:watch was a no-op (empty
-  # .changeset/ after push). No new plugin version exists → nothing to
-  # refresh. Without this guard, the orchestrator burns wall-clock and
-  # noise on a no-op /install-updates invocation every iter.
-  run grep -niE 'only.*release.*actually|when.*release.*shipped|skip.*when.*no.changeset|conditional on.*release' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-@test "work-problems P233: cache-refresh is non-blocking on /install-updates failure" {
-  # If /install-updates fails (cache miss + non-interactive fallback,
-  # marketplace fetch error, transient flake), the orchestrator MUST NOT
-  # halt the loop. Degrades to current behaviour — cache stays stale;
-  # equivalent to pre-amendment behaviour. Loop continues.
-  run grep -niE 'non.blocking.*install.updates|install.updates.*non.blocking|do not halt.*install.updates|install.updates.*failure.*continue' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-# ── Empirical evidence citation ────────────────────────────────────────────
-
-@test "work-problems P233: cache-refresh subsection cites briefing/afk-subprocess.md evidence" {
-  # The architect FLAG resolution required the evidence that subprocesses
-  # re-resolve cache on spawn (rather than inherit parent-resolved
-  # plugins). The briefing entry is dispositive — cite it inline so
-  # future readers don't re-litigate the question.
-  run grep -nE 'afk-subprocess|briefing.*subprocess.*cache' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-# ── ADR-013 Rule 5 policy authorisation ────────────────────────────────────
-
-@test "work-problems P233: cache-refresh chain rides ADR-013 Rule 5 (same authorisation as push:watch / release:watch)" {
-  # The chain MUST be silent — invocation between iters is a mechanical
-  # post-release step the framework has resolved. Same ADR-013 Rule 5
-  # citation already covers push:watch / release:watch in the within-
-  # appetite drain; the cache-refresh extends the same authorisation.
-  run grep -nE 'ADR-013 Rule 5|Rule 5 policy-authorised|policy-authorised.*ADR-013' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-# ── Mid-loop ask discipline preservation (P130) ────────────────────────────
-
-@test "work-problems P233: cache-refresh prose cites install-updates Non-interactive fallback for AskUserQuestion routing" {
-  # If install-updates' Step 5b/5c consent gate fires (cache miss /
-  # scope delta / INSTALL_UPDATES_RECONFIRM=1), the orchestrator main
-  # turn MUST NOT surface AskUserQuestion mid-loop per P130. Cite the
-  # install-updates Non-interactive fallback as the routing.
-  run grep -niE 'Non.interactive fallback|non-interactive fallback' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-@test "work-problems P233: cache-refresh prose cites ADR-044 framework-resolution boundary for AskUserQuestion-available-but-forbidden routing" {
-  # Architect advisory: cite ADR-044 alongside the Non-interactive
-  # fallback ref so future readers can trace the "AskUserQuestion-
-  # available-but-forbidden" reasoning back to the framework-resolution
-  # boundary — not just the install-updates-side dry-run mechanic.
-  run grep -nE 'ADR-044' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-# ── Composition with above-appetite branch ─────────────────────────────────
-
-@test "work-problems P233: cache-refresh fires only after within-appetite Drain action (not after above-appetite Rule 5 halt)" {
-  # Above-appetite Rule 5 halt terminates without a release; no cache
-  # refresh fires. Within-appetite convergence loops back to Drain
-  # action; cache refresh fires there. The prose must not place the
-  # chain inside the above-appetite branch.
-  run grep -niE 'within.appetite.*drain.*cache|cache.refresh.*within.appetite|after.*drain.*action' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-# ── Non-Interactive Decision Making table row ──────────────────────────────
-
-@test "work-problems P233: Decision Making table carries a post-release cache-refresh row" {
-  # The decision table is the AFK reader's quick-reference; without a
-  # row here the cache-refresh chain is buried in Step 6.5.
-  run grep -nE '\|.*[Pp]ost.release cache|\|.*[Cc]ache refresh|\|.*[Pp]lugin cache refresh' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-@test "work-problems P233: Decision Making table row cites P233" {
-  # Per ADR-037 doc-lint pattern + existing P140 / P126 / P135 sibling
-  # rows — every Decision Making table row that documents a P-driven
-  # refinement cites the driver ticket.
-  run grep -nE '\|.*cache refresh.*P233|\|.*P233.*cache refresh|\|.*post.release.*P233|\|.*P233.*post.release' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-# ── Composition cross-references ───────────────────────────────────────────
-
-@test "work-problems P233: cache-refresh prose cross-references P106 (claude plugin install no-op-when-already-installed)" {
-  # P106 is the compounding factor that makes /install-updates' explicit
-  # uninstall + install dance necessary (vs. naive claude plugin
-  # install which silent-no-ops). Cite it so future readers don't try
-  # to simplify the chain.
-  run grep -nE 'P106' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}
-
-@test "work-problems P233: cache-refresh prose cross-references P130 (orchestrator main-turn ask discipline preserved)" {
-  # P130 is the constraint the Non-interactive fallback routing exists
-  # to satisfy. Citation makes the dependency auditable.
-  run grep -nE 'P130' "$SKILL_MD"
-  [ "$status" -eq 0 ]
-}