loki-mode 11.58.0 → 11.58.2

package/SKILL.md

@@ -3,7 +3,7 @@ name: loki-mode
 description: Runs an autonomous software factory that knows what it is supposed to deliver and proves it did, with a signed receipt. Use when the user says Loki Mode or asks to build, fix or verify software autonomously.
 ---
 
-# Loki Mode v11.58.0
+# Loki Mode v11.58.2
 
 **Current entry points (Loki 10 engine):** `loki "<task>"`, `loki owner/repo#N` and `loki quick "<task>"` run the Loki 10 engine and end with a signed receipt (`loki verify`). `loki backlog owner/repo --all` runs many issues in parallel. Guide: `docs/v10/GUIDE.md`. The skill instructions below, and `loki start`, describe the previous bash loop. PRD-file starts (`loki start ./prd.md`) and the opencode provider still run on it until FC38-SWEEP ports them; the legacy-engine command and the LOKI_ENGINE switch remain removed. Prefer `loki owner/repo#N` for issues.
 
@@ -476,4 +476,4 @@ Implemented, planned and deprecated items, with the version each landed in: `ski
 
 ---
 
-**v11.58.0 | [Autonomi](https://www.autonomi.dev/) flagship product | core skill**
+**v11.58.2 | [Autonomi](https://www.autonomi.dev/) flagship product | core skill**

package/VERSION

@@ -1 +1 @@
-11.58.0
+11.58.2

package/dashboard/__init__.py

@@ -7,6 +7,6 @@ Modules:
     control: Session control helpers (status, atomic writes, origin check)
 """
 
-__version__ = "11.58.0"
+__version__ = "11.58.2"
 
 __all__ = ["__version__"]

package/docs/CLI-REFERENCE.md

@@ -1,6 +1,6 @@
 # CLI reference
 
-Generated from `loki-ts/src/cli/registry.ts` by `scripts/generate-stale-zero.sh` for v11.58.0. Do not edit by hand.
+Generated from `loki-ts/src/cli/registry.ts` by `scripts/generate-stale-zero.sh` for v11.58.2. Do not edit by hand.
 Hidden and legacy-only commands are omitted; see `docs/v10/CLI-MODERN.md` for the full inventory.
 
 | Command | Description |

package/docs/COMPETITIVE-SCORECARD.md

@@ -428,6 +428,52 @@ bash tests/test-fast-verify.sh               # prints the ms and the budget
 bash tests/test-helm-worker-scaling.sh       # renders 1, 2, 5 and 25 replicas
 ```
 
+### 3e. UX baselines re-measured on loki-mode@11.55.3, 2026-10-23 (issue #350)
+
+Tracking issue #350 recorded baselines on 11.54.0. This section re-measures what
+could be re-measured on the published 11.55.3 (main VERSION was 11.56.0). It
+extends the ledger above; it adds no parallel file. Setup: one macOS 27.0.1
+machine, node v26.5.0, npm 11.17.0, 1-minute load average about 20 (timings are
+pessimistic), loki installed with `npm i -g loki-mode@11.55.3 --prefix <run-owned
+tmp>/p2 --cache <tmp>/c2` so the user's global install was untouched. Fixtures
+were fresh one-commit git repos with `"test":"node --test"`. Every run's output
+was timestamped per line with
+`perl -MTime::HiRes=time -ne 'printf("%7.2f %s",time-$s,$_)'`. Paid spend
+(LOKI_BENCH_SPEND_APPROVED=1, founder pre-approval): one VERIFIED run, $0.05, as
+printed on its `Cost:` line.
+
+| Metric | Measured on 11.55.3 | Command or source | #350 value on 11.54.0 |
+|---|---|---|---|
+| Install | 7.65s, rc=0 | `npm i -g loki-mode@11.55.3 --prefix ... --cache ...` | 7.0s |
+| Install to first VERIFIED (trivial brief) | 7.65s install plus a 27s run (`Time: 27s`; intake 2s, wall 10s, implement 14s, verify 1s, seal 1s), $0.05 | `loki start "<brief>"` in the fixture | 31.8s total |
+| First output line | 0.71s | first timestamp of the same run | 0.56s |
+| First stage line (`[wall]`) | 6.72s | same run | 5.57s |
+| Missing provider CLI | actionable message with the fix command, rc=2, 1.40s | `PATH=/usr/bin:/bin:<dir with node and git> loki start "add x"` | rc=2 at 1.19s |
+| Flags on the core path | 0 required; `loki start --help` is 135 lines, 60 distinct `--flags`; `loki --help` lists 25 commands | `loki start --help \| wc -l`, `grep -oE -- '--[a-z][a-z0-9-]*' \| sort -u \| wc -l` | 135 lines, 60 flags, 25 commands |
+
+Observed again on 11.55.3 without a new number:
+
+- Invalid API key: the failed run ends with `Reason: exit 1 (general error)`; the
+  cause appears only in the implement log under `.loki/`, not on the terminal.
+  This matches the 11.54.0 finding. Slice: CW-16 #351.
+- A failed zero-change run leaves its empty `loki/e10-...` branch checked out
+  (`git log main..HEAD` printed 0 commits). Matches the 11.54.0 finding.
+  Slice: CW-19 #354.
+
+NOT MEASURED on 11.55.3, so the #350 11.54.0 values stand unrefreshed: crash
+recovery and `loki start --resume` (M6), false-done rate (M5, the benchmark task
+did not discriminate), and the bare `claude -p` comparison column. Do not read
+the 11.54.0 numbers as current for those rows.
+
+Honest limits on the 100x goal: install plus two model calls is the floor for
+time to first VERIFIED, so a 100x claim there is not available; 4.6x versus bare
+`claude -p` was recorded on 11.54.0 and was not re-run. The one case where the
+gap is unbounded is the invalid-key cause, which loki hides and `claude -p`
+prints. #350 stays open until its targets are met.
+
+Reproduce: the commands in the table, run inside a directory created with
+`loki_run_tmp_create`.
+
 ### 3c. Deploy surface, measured 2026-08-08
 
 Ran on the four competitor CLIs installed on this machine. Method matters here:

package/docs/INSTALLATION.md

@@ -2,7 +2,7 @@
 
 Loki Mode, by [Autonomi](https://www.autonomi.dev/), turns a spec or an issue into a verified pull request with a receipt you can check. This page covers installing it, a first run, and the options around it.
 
-**Version:** v11.58.0
+**Version:** v11.58.2
 
 **Engine note:** `loki start` is the primary command. `loki start`, `loki "<task>"`, `loki owner/repo#N` and `loki quick` all run the Loki 10 engine, whatever the flag order or input form (PRD file, issue ref, task or one-word brief). The previous loop runs only on `--engine legacy`, `LOKI_ENGINE=legacy`, or for one call when a flag Loki 10 cannot honor is present. See the [user guide](v10/GUIDE.md) and the [5-minute quickstart](QUICKSTART.md).
 

package/docs/v10/PROGRESS.md

@@ -2717,3 +2717,9 @@ v11.17.21 published (Release success on 462fbc40a; npm latest 11.17.21, next 11.
 - Blocked in review: cw04 e220a18ff (rev-a B1 signed blocks read from agent-writable run-dir files with no engine writer; B2 strict bypass; B3 sandbox fails open). Ruling: label unverified, strict absent/malformed = PARTIAL, sandbox on with malformed or non-allowlisted kind = FAILED, absent = PARTIAL until the cw01 writer lands. cw18 c1def44f5 (4 blockers incl. e10ext 1520 > 1500 lines).
 - Approved: skilllink 32890c693 (tl-324 MEDIUM, 19/0, mutant red). In review: #328 bef58ca69 (rev-c), #332 c5a6d4d2b (tl-324), #309 f230e3fce (rev-b).
 - Founder machine: global loki and ~/.claude/skills/loki-mode still at 11.17.103 (left untouched).
+
+## 2026-10-23T06:15Z (RM) checkpoint, v11.57.1..v11.58.0 shipped
+- Shipped, each with `git ls-remote origin main` equal to the local HEAD: v11.57.1 c4a7f1f4c (sdkconverge), v11.57.2 50e2dc64c (goldenfix), v11.57.3 6f1e54223 (cw19b), v11.57.4 a680eda9b (groupfix), v11.57.5 4b0bbb0ca (vscomma), v11.57.6 9556d6f16 (skilllink), v11.58.0 9e4812a14 (cw20-r2, Refs #355).
+- 05:52Z: #204 closed (test-memory-cli 25/0 on clean origin/main, log ~/loki-ci-logs/cos-204.log); RQ-07 eng-313r, RQ-08 eng-cw12, RQ-06 eng-h2h, RQ-02 eng-303c, loadaware eng-351; paid resume-cost pair GO (eng-npm, D108).
+- P0 open: publish-npm red on release.yml runs for v11.56.5 and v11.57.0 to v11.57.4 (run 39908304065 log: "dist-tags did not settle on the highest version after 8 tries"). `curl registry.npmjs.org/loki-mode/11.57.4` returned 404 at 05:58:11Z and 200 at 06:01:10Z (published 05:54:57Z), so propagation exceeded the settle loop. Dist-tags read at 06:08:16Z: latest 11.57.6, next 11.57.5. Fix owner eng-npm, HIGH review rev-a; no tag is ever written by hand.
+- Live exposure: v11.57.0 (cw01 sandbox, default LOKI_OS_SANDBOX=off) shipped before rev-b's 4 blocking findings; fix-forward cw01b and cw01c in review.