loki-mode 9.56.0 → 9.59.0

package/docs/v10/DECISIONS.md

@@ -195,3 +195,17 @@ One entry per decision: context, choice, why, how to reverse. Newest last.
 - Choice: build guards 1-3 now as slices through the normal review process (per the founder's own instruction: "each as a slice through normal review"), without interrupting or pushing anything onto the currently-frozen release train (train 1 remains frozen at `c31cb1e8` until Tests/Bun Parity/Security Audit are all green there). Guards 4-7 are process/documentation changes that also queue as slices, cut alongside 1-3, but are not blocking the reply the founder asked for (the founder's own instruction is to reply once guards 1-3 are live).
 - Why: this directive is itself direct, textbook evidence for guard 5's own thesis -- this session has repeatedly found and disclosed the SAME classes of near-miss (an agent spending disproportionate time/tokens on trivial work, a stuck background classification process, a BOARD.md edit that could have dropped rows, self-inflicted CI cancellation) as one-off narrative corrections in PROGRESS.md's drift audits, without ever converting any of them into a structural guard that makes the mistake mechanically non-repeatable. The founder's framing ("prompts drift; guards do not") is the correct fix for a pattern this session's own turn-102/120/132/138/144 audits kept re-discovering piecemeal.
 - Reverse: continue relying on per-turn discipline and drift-audit self-correction without codifying guards. Not chosen -- directly countermanded, and this session's own history is the founder's strongest evidence against it.
+
+## D27. 2026-09-27: Loki directive -- speed: a train every 20 minutes, CI replaces local gates, trains pipelined
+
+- Context: train 2 sat unpushed for more than 89 minutes. The local-ci fast tier took more than 10 minutes under swarm load (90 of 173 checks at the 600s cap, EXIT=124), and the pre-push hook's serial pytest alone exceeded 10 minutes. The first push attempt was killed by its own timeout. The second reported success through a pipe while `git ls-remote` still showed b651b98d. Loki directed: push train 2 now with PRE_PUSH_SKIP=1 (Loki-authorized; CI is the gate), never pipe git push, and verify each push by ls-remote.
+- Choice (Loki approved):
+  1. The "Local CI Before Every Push" mandate (2026-07-31) is retired. Before a push, run only syntax checks (`bash -n`, `py_compile`) plus the slice's own tests, capped at 60 seconds. GitHub CI (Tier B) is the release gate.
+  2. `.githooks/pre-push` becomes identity check + `bash -n` + red-main warning only (no pytest), target under 5 seconds; a LOW slice in train 3 (S-97).
+  3. A push is done only when `git ls-remote origin refs/heads/main` equals `git rev-parse HEAD`; recorded as evidence and enforced by a guard (S-98).
+  4. Train 3 carries the release-speed fixes first as each is approved (Part C items 1-7: S-79, S-80, S-84, S-81, the needs:gate removal inside S-84, S-82, S-83); other approved slices ride along but never delay a train.
+  5. Trains are pipelined: a train counts as released when publish-npm succeeds. The next train opens immediately, without waiting for npm's availability lag, Docker or Homebrew; train N+1 can run Tier B while train N publishes.
+  6. Cadence: push a train every 20 minutes whenever at least one merged slice exists, even a single one. New pulse violation TRAIN_LATE: more than 25 minutes since the last train push while merged-unreleased commits exist (S-98).
+- Release Manager exception, disclosed: until S-16 makes `scripts/release.sh` bump every checklist file without prompting or pushing, the Release Manager bumps VERSION with the editor tool (not a shell command) as part of the release commit. v9.57.0 (5332bfc3) was bumped this way. This is the only permitted VERSION write outside release.sh, and it ends when S-16 merges.
+- Evidence: train 2 push, `PRE_PUSH_SKIP=1 git push origin main` at 15:48:18Z: rc=0; `git rev-parse HEAD` = `git ls-remote origin refs/heads/main` = 89e350bd641a4b659faa405587a9db5bb7bcc201. Release push at 15:50:09Z: rc=0; both = 5332bfc35ccf801bdf8c67bff9fc07537ac4240f. The per-SHA concurrency from S-80 held on its first real push: Tests on 89e350bd kept running (`in_progress`) after 5332bfc3 was pushed.
+- Reverse: keep local CI as a precondition. Not chosen: it cannot finish inside the 10-minute command cap under swarm load, and it duplicates GitHub CI.

package/docs/v10/METRICS.md

@@ -123,3 +123,36 @@ ratchet-ordering deadlock: GF-4 cannot promote `P4.three-setups-resolve`
 out of pending.txt until S-19's stricter pass bar is also in the tree, or
 the merge itself becomes a ratchet regression. S-19 promoted to the
 critical path and reviewed immediately.
+
+## CI: Tests workflow shell-tests sharding (S-81, supersedes S-70)
+
+S-70 resharded shell-tests 4 -> 8 with a plain `idx % n` split and measured
+the real per-suite cost of doing so (see its commit, c6000cb9). S-81 keeps
+n=8 but replaces the index split with a deterministic greedy longest-first
+(LPT) bin-packing over measured per-suite durations, so the 8 shards are
+balanced by real cost instead of by coincidence of registration order.
+
+BEFORE numbers are measured from a real completed GitHub Actions run. AFTER
+numbers are **projected from measured per-suite durations**, not measured --
+this sandbox cannot trigger a real push, so nothing below claims to be a live
+CI wall clock for the new packing until a real run confirms it.
+
+| Metric | Value | Method |
+|---|---|---|
+| BEFORE: measured per-suite durations, source | real run 36319020918 (2026-09-27), 4 shard job logs, S-44's START/END lines | `gh run view --job <id> --log`, grepped for `END: ` |
+| BEFORE: suites with a real measured row | 496 of 496 registered suites (100%; no suite needed the default) | exact-name join against `tests/run-all-tests.sh`'s `run_test` registrations |
+| Deliberate override, not a measurement | `test-ci-json-argmax.sh` measured 1186s in that run; a parallel BACKLOG 25 slice is shrinking its fixture, so `tests/shard-durations.tsv` assumes ~60s post-fix and uses 60, not 1186 | documented in the table's own header comment |
+| BEFORE (projected from measured per-suite durations): max shard at n=4, plain `idx % 4` (current main, pre-S-70/S-81) | 451s test-time (~7.5min) | `idx % 4` replay against `tests/shard-durations.tsv`'s same measured rows (single source run, n=1) |
+| BEFORE (projected from measured per-suite durations): max shard at n=8, plain `idx % 8` (S-70's scheme) | 346s test-time (~5.8min) | same replay, `idx % 8` -- matches S-70's own commit projection (c6000cb9) of "~346s (~5.8min)" once argmax is fixed, cross-checked independently here |
+| AFTER (projected from measured per-suite durations): max shard at n=8, LPT-packed (S-81, this change) | 197s test-time (~3.3min) -- every shard lands at exactly 197s, because 1576s total / 8 divides evenly against this suite set | replay of the actual `tests/run-all-tests.sh` LPT logic via `LOKI_TEST_SHARD=i/8 LOKI_TEST_LIST=1`, summed against `tests/shard-durations.tsv` |
+| AFTER (projected from measured per-suite durations): max shard including ~29s job setup (checkout/pip/npm/bun, per S-70's measurement) | about 226s (about 3.8min), under the 6-minute target | same table + S-70's measured setup overhead |
+| Named suites land in 5 distinct shards (required) | trust-core-detect -> shard 1, review-assurance-tail -> shard 0, e2e-features -> shard 4, shellcheck -> shard 2, no-unreachable -> shard 3 | same LPT replay |
+| Partition proof | `tests/test-shard-coverage.sh`: n in [2, 3, 4, 6, 8] each give union=496, 0 duplicates, 0 missing, no empty shard -- driven through the real runtime path (`LOKI_TEST_LIST=1`), not a second copy of the packing algorithm | `bash tests/test-shard-coverage.sh` |
+
+**What still needs real-CI verification** (cannot be done from this sandbox):
+an actual GitHub Actions run of the updated `test.yml` at n=8 with the new
+packing, to confirm the real wall clock lands near the ~3.8min projection
+rather than something job-setup variance or GitHub's own queueing missed, and
+that `test-ci-json-argmax.sh` actually lands near 60s once its BACKLOG 25
+fixture fix ships (until then this table's 60s row is an assumption, not a
+measurement, as stated in `tests/shard-durations.tsv`'s own header).

package/docs/v10/OPERATING-MODEL.md

@@ -0,0 +1,47 @@
+# Loki company operating model (always loaded; never exceed 90 lines)
+
+The CEO is Lokesh Mure (Loki), the founder. Everyone else is an agent. The company ships Loki Mode v10 per docs/V10-VISION.md. Order when goals conflict: moat > Seal accuracy > delivered accuracy > cost > speed. Velocity target: 30-60 releases per day.
+
+## Roles and models (pin every agent's model explicitly, D13)
+- Chief of Staff (the main session, opus): runs the operating rhythm, owns BOARD.md dispatch, keeps every seat busy, escalates only to Loki, never writes product code.
+- CTO (opus, advisor): architecture and quality calls, approves any change to the moat, gates or release policy.
+- Project Managers (opus, advisors): flow, bottlenecks, staffing. Each PM owns one workstream and its board rows.
+- Architects (opus): split every large task into 10-40 independent slices with non-overlapping file sets and Wall checks. They design; they do not implement.
+- Product Owners (sonnet): backlog ranking, slice cards (60 lines max: goal, file set, Wall checks, commands, budget, tier).
+- Tech Leads (sonnet): LOW and MEDIUM reviews, unanimous (D12).
+- Reviewers for HIGH-tier work (opus): moat, verifier, Seal, auth, security. Unanimous; a reproduced blocking finding always blocks (D12).
+- Software Engineers (sonnet): one slice each, worktree isolation, single writer, red-then-green tests.
+- Release Manager (sonnet, exactly one): the only role that pushes main, bumps VERSION, tags or publishes.
+
+## Leadership lock (one company per repo)
+- The leader holds .loki/v10-leader (PID and start time). A session becomes leader only if that file is missing or its PID is dead (kill -0 fails).
+- Non-leader sessions act as extra engineers or reviewers on BOARD.md slices, or answer Loki. They never push, bump VERSION, release or rewrite BOARD.md.
+
+## Nobody sits idle
+- The Chief of Staff keeps 8-16 engineers busy whenever ready slices exist. When anyone finishes, the next slice is assigned in the same turn.
+- The ready queue never drops below 2x the number of engineers; PMs and POs refill it.
+- Idle seats while work exists, or an empty ready queue, is a P0 violation.
+
+## Time limits (anything slower gets redesigned, not waited on)
+- Budgets: trivial ops use scripts, never agents (target under 10s). LOW engineer 15 min, MEDIUM 30 min. Reviewers 30 min, HIGH adversarial reviewer 60 min.
+- Any shell command: timeout 10 min max; target under 2 min.
+- An agent or command over budget is stopped, recorded in METRICS.md, and re-sliced smaller.
+
+## Release pipeline
+- Tier A fast gate on every change (target 60-120s): syntax checks, changed-file shellcheck, diff-selected tests, moat suite split per property. It is feedback, not release authority.
+- Tier B full verification on every main commit (target 5-6 min, hard 10 min per job, never cancelled on main).
+- Release = a lookup: publish a commit whose tree already has a Tier B pass. A VERSION-only bump reuses its parent's verdict. Target: 2 min from verified to npm.
+- Trains: the Release Manager batches merged slices; one push per train; releases when verified; the next train opens immediately. Never hold a train for one HIGH slice.
+- If Tier B fails on main: stop releases, open a P0 fix-forward slice. Never unpublish; npm deprecate is allowed.
+
+## Large tasks (for example "rebuild the dashboard")
+- The Architect splits the task into 10-40 slices behind a feature flag.
+- Engineers build them in parallel; each merges and releases behind the flag as soon as it is green.
+- The flag flips when the Wall checks for the whole feature pass.
+- Releases never pause for a big task.
+
+## Discipline
+- Evidence or it did not happen: every claim cites a command plus its exit code or output line.
+- Every incident becomes a guard with a test (docs/v10/GUARDS.md, D26).
+- The pulse block (scripts/v10-pulse.sh) is injected every turn. Act on the top VIOLATION first.
+- Never kill by name or pattern, force-push, git add -A, or bypass branch protection.

package/docs/v10/PROGRESS.md

@@ -1035,3 +1035,176 @@ else should take priority over.
     mean the SAME OS family as the failing CI job, not just "my own
     machine" -- a macOS-only local repro is not sufficient evidence to
     rule out a Linux-specific regression, ever.
+
+- 2026-09-27T14:13Z: Tests rerun of shard 0 on `c31cb1e8` completed
+  success (`gh run view 36321857199 --json status,conclusion` ->
+  `completed success`). Train 1 verified: Tests and Bun Parity green on
+  the frozen SHA.
+- 2026-09-27T14:16:24Z: pushed release commit `b651b98d` (v9.56.0,
+  `c31cb1e8..b651b98d`). Only version strings, CHANGELOG and a fresh
+  deterministic dist build on top of docs-only commits. Release run
+  36325316189, Tests 36325316218, Bun Parity 36325316184, Security
+  Audit 36325316176 all started on `b651b98d`. Main frozen until publish.
+- Found while releasing: the committed loki-ts/dist was not a fresh
+  build of src (a second local build was byte-identical to the first,
+  so the build is deterministic; after normalizing minified identifiers
+  the only semantic difference is the version string). Shipped the
+  fresh build.
+
+## Drift audit, turn 174
+
+**Matched:** the CEO directive's Part A shipped first: train 1 was
+verified on its frozen SHA and released as v9.56.0 before any Part C
+work merged. Leader lock taken (`.loki/v10-leader`, PID 74619). 6
+engineers and 10 reviewers dispatched in parallel on the named items;
+pulse `Active builder worktrees` rose from 5 to 19, clearing
+IDLE_BUILDERS. Trivial ops (committing finished builds, BOARD edits,
+version bump) were done directly, not by agents (D26 guard 2).
+
+**Drifted, self-caught:** two builders (S-74, S-82) stopped short of
+committing and asked for approval, following the global
+approve-before-commit rule the repo's standing authorization waives.
+Fixed by committing their work directly; S-78 (CLAUDE.md trim) is the
+real fix, since it moves the operating model into every session.
+The release also surfaced a stale committed loki-ts/dist that no gate
+caught on train 1's own commit; the fast-tier dist-freshness check
+exists but the train skipped local-ci. Correction: run
+`bash scripts/local-ci.sh` (fast tier) on every train's freeze commit
+before pushing it, not only on release commits.
+
+## Drift audit, turn 180
+
+**Near-miss caught, becomes a GUARDS.md entry (S-76):** S-72's
+worktree cleanup used "directory modified in the last 90 minutes" as
+its liveness signal. Its own `git -C <wt> status` calls refresh each
+worktree's index and touch mtimes (pulse jumped to "Active builder
+worktrees: 123 of 129"), so the signal went dead, and a live agent
+that has not yet committed or written a file looks safe by every
+other rule. Caught before any removal (agent reported "0 worktrees
+removed so far"). Fix applied: an explicit exclusion list of live and
+approved-unmerged worktrees, plus `.git` file birth time (`stat -f %B`)
+as the recency signal. Guard to build: the worktree-cleanup tool must
+take the live-agent list as input and must not use mtime.
+
+**Matched:** trains stay frozen while v9.56.0 waits on Tests at
+b651b98d; all wave-2 slices build in parallel; every review that
+found a blocker went back to the same engineer with the reviewers'
+exact reproductions (S-73, S-74, S-75), not to a fresh agent.
+
+## Drift audit, turn 186
+
+**Matched:** three reworks (S-74 round 2, S-75, S-18 round 3) came
+back within their budgets and each went straight to re-review with the
+prior reviewers' exact reproductions as the checklist. Reviews are
+finding real defects (S-73 x2 REJECT, S-74 CONCERN x2, S-75 REJECT),
+which is the review gate working, not noise.
+
+**Drifted, recurring:** a third builder (S-85) stopped short of
+committing to "wait for approval" (S-74, S-82 did the same). Root cause
+is the old CLAUDE.md commit workflow text that every subagent loads;
+S-78's trim removes it. Until S-78 merges, slice cards should say
+"commit when done, you are authorized" explicitly.
+
+## Drift audit, turn 192
+
+**Matched:** release train 1 is one job from publishing (only Tests
+shard 0, the argmax shard, still running on b651b98d); the two slices
+that remove that shard's cost (S-79 fixture shrink, S-81 balanced
+sharding) are built or nearly built. S-72 freed 12 GB with zero forced
+removals.
+
+**Watch item:** a builder reported an unusually clean number (all 8
+shards projected at exactly 197 s). Reviewers are instructed to
+recompute it from the raw table rather than trust the report. Clean
+numbers get the same evidence bar as alarming ones.
+
+## Drift audit, turn 198
+
+**Recurrence (guard review rule, D26 item 7):** the "temp file then
+replace narrows mode 644 -> 600" defect appeared twice this session:
+S-16 (scripts/release.sh, mktemp + mv) and now S-74 (v10-ops.sh,
+mkstemp + os.replace). Both were caught only by review. A recurrence
+means the fix-per-slice approach is wrong; the guard belongs in
+GUARDS.md (S-76): a shared helper for atomic in-place rewrite that
+copies mode and resolves symlinks, plus a fixture that asserts mode is
+preserved, and a lint that flags bare mkstemp/mktemp followed by a
+replace without a mode copy.
+
+**Second recurring class:** locating a table cell by value shape
+instead of by header (S-74's Status-cell finder). The pulse's own
+BOARD parser (S-75) uses a position-independent shape scan too;
+S-75's reviewers found it safe because it requires a known token plus
+a full timestamp. Same rule should be written down once for every
+BOARD reader and writer.
+
+**Matched:** release still gated only on Tests shard 0 at b651b98d;
+every finished build this turn went straight to review; three
+engineers again stopped short of committing, committed directly.
+
+- 2026-09-27T14:42Z: Tests on b651b98d completed success (run
+  36325316218); Bun Parity (36325316184) and Security Audit
+  (36325316176) also success. required-ci passed.
+- 2026-09-27T14:43:50Z: GitHub Release v9.56.0 published
+  (`gh release view v9.56.0` publishedAt).
+- 2026-09-27T14:44:53Z: publish-npm job step "Publish to npm" success
+  (`gh api .../jobs/108641497036`). At 14:45:06Z the registry still
+  returned 404 for loki-mode@9.56.0: propagation window, re-checking.
+  Docker publish still running.
+
+## Drift audit, turn 204
+
+**Train 1 measured:** freeze push 13:17:07Z -> Tests green 14:13Z (one
+flaky-shard rerun) -> release commit push 14:16:24Z -> gates green
+14:42Z -> GitHub Release 14:43:50Z -> npm publish step 14:44:53Z. About
+26 min from release push to publish, almost all of it Tests on the
+release SHA re-running what c31cb1e8 had already verified. That is
+exactly the cost S-84 (verdict reuse) removes.
+
+**Matched:** the release did not wait for any HIGH slice (S-18 is in
+round 4 and did not block, per D25).
+
+## Drift audit, turn 210
+
+**npm propagation, not assumed:** 5 minutes after the publish-npm step
+("npm publish --access public", exit 0 at 14:44:53Z) the registry still
+returns 404 for loki-mode@9.56.0 (curl registry.npmjs.org, bypassing
+the npm client cache; dist-tag latest 9.55.0). npm publish only exits 0
+after the registry accepts the package, so this is CDN or packument
+caching; memory records one prior case where six agreeing probes read a
+stale packument. Not republishing (the version would be rejected as a
+duplicate anyway). Re-checking; if still absent at +15 min, investigate
+with npm support channels, never republish or bump.
+
+**Reviews keep finding real defects in fixes-to-fixes:** S-74 round 3
+fixed two bugs and introduced a third (short rows refused), caught
+because the reviewer measured the real board's row shapes instead of
+trusting a first-row-only test. Rule for slice cards touching BOARD.md
+tooling: test against every row of a copy of the real board, not a
+sample.
+
+## Turn 216 drift audit (2026-09-27T14:58Z)
+- Train 1 npm: publish-npm logged `+ loki-mode@9.56.0` and "being processed" at 14:44:51Z; registry still 404 at 14:57 (13 min). Not republishing; per the stale-packument lesson, repeated reads share one CDN cache and are not independent. publish-docker still in_progress (run 36325316189).
+- Pulse NO_RECENT_RELEASE reads npm, so it stays red until 9.56.0 is visible there. That is correct: npm is the channel users install from.
+- Drift caught: two MEDIUM builders (S-87, S-91) ran past 30 min without check-in. The newly merged S-75 pulse flagged them within minutes of landing. Guard 3 works.
+- Train 2 assembled on main: S-71, S-75, S-78, S-80, S-86, S-90 (8 commits). S-81 is held for S-79 (same runner file, and the argmax suite duration decides its packing); S-82 is held for shard times under 10 min.
+
+## Train 1 shipped to npm (2026-09-27T15:00Z)
+- npm: `curl registry.npmjs.org/loki-mode` gives dist-tags latest 9.56.0, time["9.56.0"] = 2026-09-27T15:00:06.452Z. publish-npm reported acceptance at 14:44:51Z, so npm took 15m15s to make it available. The 404s in between were npm's processing delay, not a failed publish; not republishing was correct.
+- publish-docker is still in_progress at 15:08 (run 36325316189).
+- Measurement for the Part C release-lookup target (2 min verified-to-npm): GitHub Release to npm availability was 16m16s, of which about 15 min is npm-side processing we do not control. The S-84 reuse only shortens required-ci.
+
+## Turn 240 drift audit (2026-09-27T15:24Z)
+- Near-miss, false green: the train 2 background push "completed (exit code 0)", but `git ls-remote origin refs/heads/main` still returned b651b98d. `timeout 600 git push ... | tail -5` reported tail's status, not the push's. The first attempt had already been killed by a 120s timeout during the pre-push hook's serial pytest. Correction: after every push, compare `git ls-remote` with `git rev-parse HEAD`; never trust a piped exit code. GUARDS.md candidate (S-76 follow-up): a push helper in scripts/v10-ops.sh that asserts remote == local.
+- The local-ci fast tier took more than 10 min under swarm load (90 of 173 checks at 600s, EXIT=124). Its one real failure (CLAUDE.md tool count after S-78) was fixed in 6f1e68eb. This conflicts with the 10-min command cap; S-91 (Tier A) is the planned replacement. Until then, release on GitHub CI plus the packaging checks run by hand (npm pack contents: 938 entries, 9/9 required).
+- Standing drift check: slices finished inside workflows leave BOARD rows at building until the orchestrator reads the result, which fed 5 false AGENT_OVER_BUDGET hits at 15:15. Read workflow results the same turn they complete.
+
+## Trains 2 and 3 under D27 (2026-09-27)
+| Time (UTC) | Event | Evidence |
+|---|---|---|
+| 15:36 | Train 2 assembled (12 slices); third push attempt still in the pre-push serial pytest | push-train2 log |
+| 15:47 | Loki directive D27 received; background push stopped (TaskStop bw1fgezxd) | |
+| 15:48:18-21 | Train 2 pushed: PRE_PUSH_SKIP=1 git push origin main, rc=0 | HEAD = ls-remote = 89e350bd641a4b659faa405587a9db5bb7bcc201 |
+| 15:50:09-12 | Train 2 release commit v9.57.0 pushed, rc=0 | HEAD = ls-remote = 5332bfc35ccf801bdf8c67bff9fc07537ac4240f |
+| 15:50 | S-80 verified on a real push: Tests on 89e350bd stayed in_progress after 5332bfc3 landed | gh run list |
+| 15:53:23-26 | Train 3 (S-79, S-81, S-82, S-83, S-85) plus release commit v9.58.0 pushed, rc=0 | HEAD = ls-remote = c2eccb21c1273b27a385aa785fedee3269c4f605 |
+- Next: each train counts as released when its publish-npm job succeeds. Train 4 opens now: S-84 (Part C items 3 and 5) once approved, plus S-97 and S-98 (D27 hook and cadence guard). Target push by 16:13 (20-minute cadence).

package/docs/v10/SWARM.md

@@ -223,6 +223,30 @@ queue. Never unpublishes or deletes a version.
   agent-authored slices never combine untrusted input, an agent, and push
   rights either.
 
+## Slice card template
+
+Binding: the orchestrator dispatches every builder and reviewer from a slice
+card below, never from conversation history or its own recall of prior
+turns -- the card is the sole source of scope for that agent. While the
+swarm runs, the orchestrator's own live view stays limited to
+`docs/v10/CONTROL.md`, the pulse block, and `docs/v10/BOARD.md`; it does not
+re-derive scope from scrollback.
+
+<!-- SLICE-CARD-TEMPLATE:START -->
+Slice: S-<id>
+Goal: <one sentence, the acceptance condition for this slice>
+Tier: LOW | MEDIUM | HIGH
+Model: <pinned model, e.g. opus, sonnet -- never inherited, see D13>
+Budget: <minutes>
+Files:
+  - <path>
+  - <path>
+Wall checks:
+  - <command or assertion that must pass before review>
+Commands:
+  - <exact shell command(s) the builder runs>
+<!-- SLICE-CARD-TEMPLATE:END -->
+
 ## Metrics (appended to METRICS.md after every wave)
 
 - releases in the last 24h

package/loki-ts/dist/loki.js

@@ -1,5 +1,5 @@
 // @bun
-var ct=Object.create;var{getPrototypeOf:lt,defineProperty:OG,getOwnPropertyNames:it}=Object;var at=Object.prototype.hasOwnProperty;function ot($){return this[$]}var st,nt,rt=($,X,Q)=>{var z=$!=null&&typeof $==="object";if(z){var Z=X?st??=new WeakMap:nt??=new WeakMap,K=Z.get($);if(K)return K}Q=$!=null?ct(lt($)):{};let J=X||!$||!$.__esModule?OG(Q,"default",{value:$,enumerable:!0}):Q;for(let q of it($))if(!at.call(J,q))OG(J,q,{get:ot.bind($,q),enumerable:!0});if(z)Z.set($,J);return J};var Zq=($,X)=>()=>(X||$((X={exports:{}}).exports,X),X.exports);var tt=($)=>$;function et($,X){this[$]=tt.bind(null,X)}var B1=($,X)=>{for(var Q in X)OG($,Q,{get:X[Q],enumerable:!0,configurable:!0,set:et.bind(X,Q)})};var s=($,X)=>()=>($&&(X=$($=0)),X);var w5=import.meta.require;var HR={};B1(HR,{lokiDir:()=>h0,homeLokiDir:()=>GQ,findRepoRootForVersion:()=>LG,REPO_ROOT:()=>O1});import{resolve as h2,dirname as jG}from"path";import{fileURLToPath as $e}from"url";import{existsSync as Kq}from"fs";import{homedir as Xe}from"os";function Qe(){let $=UR;for(let X=0;X<6;X++){if(Kq(h2($,"VERSION"))&&Kq(h2($,"autonomy/run.sh")))return $;let Q=jG($);if(Q===$)break;$=Q}return h2(UR,"..","..","..")}function LG($){let X=$;for(let Q=0;Q<6;Q++){if(Kq(h2(X,"VERSION"))&&Kq(h2(X,"autonomy/run.sh")))return X;let z=jG(X);if(z===X)break;X=z}return h2($,"..","..","..")}function h0(){return process.env.LOKI_DIR??h2(process.cwd(),".loki")}function GQ(){return h2(Xe(),".loki")}var UR,O1;var k1=s(()=>{UR=jG($e(import.meta.url));O1=Qe()});import{readFileSync as ze}from"fs";import{resolve as Ze,dirname as Ke}from"path";import{fileURLToPath as Je}from"url";function L9(){if(t3!==null)return t3;let $="9.56.0";if(typeof $==="string"&&$.length>0)return t3=$,t3;try{let X=Ke(Je(import.meta.url)),Q=LG(X);t3=ze(Ze(Q,"VERSION"),"utf-8").trim()}catch{t3="unknown"}return t3}var t3=null;var Jq=s(()=>{k1()});var jR={};B1(jR,{runOrThrow:()=>Te,run:()=>X1,readStreamCapped:()=>qq,commandVersion:()=>Fe,commandExists:()=>g5,ShellError:()=>CG,MAX_STDOUT_BYTES:()=>OR});async function qq($,X=OR){let Q=$.getReader(),z=new TextDecoder,Z="",K=0;try{while(K<X){let{done:J,value:q}=await Q.read();if(J)break;if(!q)continue;if(K+=q.byteLength,K>X){let V=q.byteLength-(K-X);Z+=z.decode(q.subarray(0,V),{stream:!0});break}Z+=z.decode(q,{stream:!0})}Z+=z.decode()}finally{try{await Q.cancel()}catch{}Q.releaseLock()}return Z}async function X1($,X={}){let Q=Bun.spawn({cmd:[...$],stdout:"pipe",stderr:"pipe",env:X.env?{...process.env,...X.env}:process.env,cwd:X.cwd}),z,Z;if(X.timeoutMs&&X.timeoutMs>0)z=setTimeout(()=>{try{Q.kill("SIGTERM")}catch{}Z=setTimeout(()=>{try{Q.kill("SIGKILL")}catch{}},2000)},X.timeoutMs);try{let[K,J,q]=await Promise.all([qq(Q.stdout),new Response(Q.stderr).text(),Q.exited]);return{stdout:K,stderr:J,exitCode:q}}finally{if(z)clearTimeout(z);if(Z)clearTimeout(Z)}}async function Te($,X={}){let Q=await X1($,X);if(Q.exitCode!==0)throw new CG(`command failed (${Q.exitCode}): ${$.join(" ")}`,Q.exitCode,Q.stdout,Q.stderr);return Q}async function g5($){let X=De($),Q=await X1(["sh","-c",`command -v ${X}`],{timeoutMs:5000});if(Q.exitCode===0)return Q.stdout.trim()||null;return null}function De($){if(!/^[A-Za-z0-9._/-]+$/.test($))throw Error(`refused to shell-escape suspect token: ${$}`);return $}async function Fe($,X="--version"){if(!await g5($))return null;let z=await X1([$,X],{timeoutMs:5000});if(z.exitCode!==0)return null;return((z.stdout||z.stderr).split(/\r?\n/)[0]?.trim()??"")||null}var OR=16777216,CG;var y8=s(()=>{CG=class CG extends Error{message;exitCode;stdout;stderr;constructor($,X,Q,z){super($);this.message=$;this.exitCode=X;this.stdout=Q;this.stderr=z;this.name="ShellError"}}});function g2($){return Re?"":$}var Re,p0,$5,Z1,y61,A1,f1,m5,n;var e7=s(()=>{Re=(process.env.NO_COLOR??"").length>0;p0=g2("\x1B[0;31m"),$5=g2("\x1B[0;32m"),Z1=g2("\x1B[1;33m"),y61=g2("\x1B[0;34m"),A1=g2("\x1B[0;36m"),f1=g2("\x1B[1m"),m5=g2("\x1B[2m"),n=g2("\x1B[0m")});import{existsSync as fe}from"fs";async function Z2(){if(BQ!==void 0)return BQ;let $="/opt/homebrew/bin/python3.12";if(fe($))return BQ=$,$;let X=await g5("python3.12");if(X)return BQ=X,X;let Q=await g5("python3");return BQ=Q,Q}async function I4($,X={}){let Q=await Z2();if(!Q)return{stdout:"",stderr:"python3 not found",exitCode:127};return X1([Q,"-c",$],X)}var BQ;var m2=s(()=>{y8()});var hR={};B1(hR,{runStatus:()=>J00});import{existsSync as u5,readFileSync as T9,readdirSync as xR,statSync as kR}from"fs";import{resolve as A5,basename as re}from"path";import{homedir as te}from"os";function SR($){let X=Math.trunc($);if(X>=1e6)return`${(Math.trunc(X/1e6*10)/10).toFixed(1)}M`;if(X>=1000)return`${(Math.trunc(X/1000*10)/10).toFixed(1)}K`;return String(X)}function yR($,X,Q){if(X===0)return null;let z=Math.trunc($*100/X),Z=Math.trunc($*Yq/X);if(Z>Yq)Z=Yq;let K=Yq-Z,J=$5;if(z>=80)J=p0;else if(z>=50)J=Z1;let q="=".repeat(Math.max(0,Z))+" ".repeat(Math.max(0,K)),V=SR($),Y=SR(X);return`  ${f1}${Q}${n} ${J}[${q}]${n} ${z}% (${V} / ${Y})`}async function $00(){if(await g5("jq"))return!0;return process.stdout.write(`${p0}Error: jq is required but not installed.${n}
+var ct=Object.create;var{getPrototypeOf:lt,defineProperty:OG,getOwnPropertyNames:it}=Object;var at=Object.prototype.hasOwnProperty;function ot($){return this[$]}var st,nt,rt=($,X,Q)=>{var z=$!=null&&typeof $==="object";if(z){var Z=X?st??=new WeakMap:nt??=new WeakMap,K=Z.get($);if(K)return K}Q=$!=null?ct(lt($)):{};let J=X||!$||!$.__esModule?OG(Q,"default",{value:$,enumerable:!0}):Q;for(let q of it($))if(!at.call(J,q))OG(J,q,{get:ot.bind($,q),enumerable:!0});if(z)Z.set($,J);return J};var Zq=($,X)=>()=>(X||$((X={exports:{}}).exports,X),X.exports);var tt=($)=>$;function et($,X){this[$]=tt.bind(null,X)}var B1=($,X)=>{for(var Q in X)OG($,Q,{get:X[Q],enumerable:!0,configurable:!0,set:et.bind(X,Q)})};var s=($,X)=>()=>($&&(X=$($=0)),X);var w5=import.meta.require;var HR={};B1(HR,{lokiDir:()=>h0,homeLokiDir:()=>GQ,findRepoRootForVersion:()=>LG,REPO_ROOT:()=>O1});import{resolve as h2,dirname as jG}from"path";import{fileURLToPath as $e}from"url";import{existsSync as Kq}from"fs";import{homedir as Xe}from"os";function Qe(){let $=UR;for(let X=0;X<6;X++){if(Kq(h2($,"VERSION"))&&Kq(h2($,"autonomy/run.sh")))return $;let Q=jG($);if(Q===$)break;$=Q}return h2(UR,"..","..","..")}function LG($){let X=$;for(let Q=0;Q<6;Q++){if(Kq(h2(X,"VERSION"))&&Kq(h2(X,"autonomy/run.sh")))return X;let z=jG(X);if(z===X)break;X=z}return h2($,"..","..","..")}function h0(){return process.env.LOKI_DIR??h2(process.cwd(),".loki")}function GQ(){return h2(Xe(),".loki")}var UR,O1;var k1=s(()=>{UR=jG($e(import.meta.url));O1=Qe()});import{readFileSync as ze}from"fs";import{resolve as Ze,dirname as Ke}from"path";import{fileURLToPath as Je}from"url";function L9(){if(t3!==null)return t3;let $="9.59.0";if(typeof $==="string"&&$.length>0)return t3=$,t3;try{let X=Ke(Je(import.meta.url)),Q=LG(X);t3=ze(Ze(Q,"VERSION"),"utf-8").trim()}catch{t3="unknown"}return t3}var t3=null;var Jq=s(()=>{k1()});var jR={};B1(jR,{runOrThrow:()=>Te,run:()=>X1,readStreamCapped:()=>qq,commandVersion:()=>Fe,commandExists:()=>g5,ShellError:()=>CG,MAX_STDOUT_BYTES:()=>OR});async function qq($,X=OR){let Q=$.getReader(),z=new TextDecoder,Z="",K=0;try{while(K<X){let{done:J,value:q}=await Q.read();if(J)break;if(!q)continue;if(K+=q.byteLength,K>X){let V=q.byteLength-(K-X);Z+=z.decode(q.subarray(0,V),{stream:!0});break}Z+=z.decode(q,{stream:!0})}Z+=z.decode()}finally{try{await Q.cancel()}catch{}Q.releaseLock()}return Z}async function X1($,X={}){let Q=Bun.spawn({cmd:[...$],stdout:"pipe",stderr:"pipe",env:X.env?{...process.env,...X.env}:process.env,cwd:X.cwd}),z,Z;if(X.timeoutMs&&X.timeoutMs>0)z=setTimeout(()=>{try{Q.kill("SIGTERM")}catch{}Z=setTimeout(()=>{try{Q.kill("SIGKILL")}catch{}},2000)},X.timeoutMs);try{let[K,J,q]=await Promise.all([qq(Q.stdout),new Response(Q.stderr).text(),Q.exited]);return{stdout:K,stderr:J,exitCode:q}}finally{if(z)clearTimeout(z);if(Z)clearTimeout(Z)}}async function Te($,X={}){let Q=await X1($,X);if(Q.exitCode!==0)throw new CG(`command failed (${Q.exitCode}): ${$.join(" ")}`,Q.exitCode,Q.stdout,Q.stderr);return Q}async function g5($){let X=De($),Q=await X1(["sh","-c",`command -v ${X}`],{timeoutMs:5000});if(Q.exitCode===0)return Q.stdout.trim()||null;return null}function De($){if(!/^[A-Za-z0-9._/-]+$/.test($))throw Error(`refused to shell-escape suspect token: ${$}`);return $}async function Fe($,X="--version"){if(!await g5($))return null;let z=await X1([$,X],{timeoutMs:5000});if(z.exitCode!==0)return null;return((z.stdout||z.stderr).split(/\r?\n/)[0]?.trim()??"")||null}var OR=16777216,CG;var y8=s(()=>{CG=class CG extends Error{message;exitCode;stdout;stderr;constructor($,X,Q,z){super($);this.message=$;this.exitCode=X;this.stdout=Q;this.stderr=z;this.name="ShellError"}}});function g2($){return Re?"":$}var Re,p0,$5,Z1,y61,A1,f1,m5,n;var e7=s(()=>{Re=(process.env.NO_COLOR??"").length>0;p0=g2("\x1B[0;31m"),$5=g2("\x1B[0;32m"),Z1=g2("\x1B[1;33m"),y61=g2("\x1B[0;34m"),A1=g2("\x1B[0;36m"),f1=g2("\x1B[1m"),m5=g2("\x1B[2m"),n=g2("\x1B[0m")});import{existsSync as fe}from"fs";async function Z2(){if(BQ!==void 0)return BQ;let $="/opt/homebrew/bin/python3.12";if(fe($))return BQ=$,$;let X=await g5("python3.12");if(X)return BQ=X,X;let Q=await g5("python3");return BQ=Q,Q}async function I4($,X={}){let Q=await Z2();if(!Q)return{stdout:"",stderr:"python3 not found",exitCode:127};return X1([Q,"-c",$],X)}var BQ;var m2=s(()=>{y8()});var hR={};B1(hR,{runStatus:()=>J00});import{existsSync as u5,readFileSync as T9,readdirSync as xR,statSync as kR}from"fs";import{resolve as A5,basename as re}from"path";import{homedir as te}from"os";function SR($){let X=Math.trunc($);if(X>=1e6)return`${(Math.trunc(X/1e6*10)/10).toFixed(1)}M`;if(X>=1000)return`${(Math.trunc(X/1000*10)/10).toFixed(1)}K`;return String(X)}function yR($,X,Q){if(X===0)return null;let z=Math.trunc($*100/X),Z=Math.trunc($*Yq/X);if(Z>Yq)Z=Yq;let K=Yq-Z,J=$5;if(z>=80)J=p0;else if(z>=50)J=Z1;let q="=".repeat(Math.max(0,Z))+" ".repeat(Math.max(0,K)),V=SR($),Y=SR(X);return`  ${f1}${Q}${n} ${J}[${q}]${n} ${z}% (${V} / ${Y})`}async function $00(){if(await g5("jq"))return!0;return process.stdout.write(`${p0}Error: jq is required but not installed.${n}
 `),process.stdout.write(`Install with:
 `),process.stdout.write(`  brew install jq    (macOS)
 `),process.stdout.write(`  apt install jq     (Debian/Ubuntu)
@@ -1349,4 +1349,4 @@ Set LOKI_LEGACY_BASH=1 to force the bash CLI for every command.
 `),2}case"start":{let{runStart:z}=await Promise.resolve().then(() => (ut(),mt));return z(Q)}default:return process.stderr.write(`Unknown command: ${X}
 `),process.stderr.write(dt),2}}PR();process.on("SIGINT",()=>process.exit(130));process.on("SIGTERM",()=>process.exit(143));var N61=await B61(Bun.argv.slice(2));process.exit(N61);
 
-//# debugId=AF4E056DAC4D3D5B742E33E3BB5AEAB7
+//# debugId=BE3483EEAD4BDCD4D66273ADAB0A2BC0

package/mcp/__init__.py

@@ -75,4 +75,4 @@ try:
 except ImportError:
     __all__ = ['mcp']
 
-__version__ = '9.56.0'
+__version__ = '9.59.0'

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "loki-mode",
   "mcpName": "io.github.asklokesh/loki-mode",
-  "version": "9.56.0",
+  "version": "9.59.0",
   "description": "Loki Mode by Autonomi. Autonomous spec-to-product system: takes a PRD, GitHub issue, OpenAPI/JSON/YAML, or one-line brief to a deployed app via the RARV-C closure loop with 8 quality gates. Provider-agnostic (Claude Code, OpenAI Codex, Cline, Aider, opencode).",
   "keywords": [
     "agent",

package/plugins/loki-mode/.claude-plugin/plugin.json

@@ -2,7 +2,7 @@
   "$schema": "https://json.schemastore.org/claude-code-plugin-manifest.json",
   "name": "loki-mode",
   "displayName": "Loki Mode",
-  "version": "9.56.0",
+  "version": "9.59.0",
   "description": "Autonomous spec-to-product build system with a built-in trust layer (RARV-C closure loop, 8 quality gates, completion council). Ships Loki's spec-hardening, drift-detection, and deterministic PR verification commands plus the Loki MCP server.",
   "author": {
     "name": "Autonomi",