Skip to content

v5.18.0

Choose a tag to compare

@Kanevry Kanevry released this 11 Nov 18:05
· 4 commits to main since this release

Added

  • session-start lays native liveness over the peer banner (#1052): on Claude Code the coordinator calls ListAgents once in Phase 0.5 and marks each registry peer busy/idle and reachable by SendMessage, joined through ~/.claude/sessions/*.json on the session id the host registry also keys on, never on cwd (two peers in one working copy share it) or a name prefix. Display only; harnesses without the tool print that liveness is not available.

  • Lead-guided runs for grill, ux-grill, deepen, test-audit and discovery (#1976): a session that nobody answers — lead-guided by a fleet coordinator, or headless (claude -p, codex exec) — no longer stalls or skips these skills. skills/_shared/lead-guided-mode.md names the four detection signals, the three duties (take the own recommendation, park owner-class questions, record one Entscheidung (autonom, Lead-geführt): … line per answer) and the owner classes (external effect incl. a public tracker, money, accounts, deletion, security, critical product decision). grill runs a "Selbst-Grill" that answers only with cited evidence; deepen executes the rewrite and names the original as the rejected option; ux-grill, test-audit and discovery take their recommended dispositions; a lead-guided run ends with a report whose open points are statements, never a question. Interactive behaviour is unchanged. grill and deepen carry the text themselves so the OpenAI skills-only export stays self-contained (export check: 0 violations). Measured with headless probes on 2026-11-08: deepen and ux-grill made 0 AskUserQuestion calls, wrote their decision lines and ran to the end.

  • Viewport pair per page and /test --target-kind (#1962): the playwright-driver capture recipe writes ax-snapshots/<page-slug>-w1440|w390.yaml plus screenshots for every page (rubric-v2 convention, locator.ariaSnapshot()) and a visited-urls.json with the URL actually driven per page and viewport. /test --target-kind live|production-build|dev-server --target-url <url> requests a kind; LIVE RUN / PRODUCTION-BUILD RUN is granted only when the visited URLs have the target's origin, otherwise the report says dev-server and names the mismatch. checkReportComplete receives routes, never slugs; the evaluator is the single source of the report header line.

  • /plan new applies the research budget with a fixed time limit (#1922): each research agent checks a 4-minute wall-clock limit itself, because /plan waits in the foreground and the coordinator cannot stop a running agent; the call limits of the /plan budget stay. Headless and autonomous runs cannot raise the limits, every research prompt carries them, and PRD Section 2 gets a "Research Budget" subsection with allowed and used calls per agent. skills/plan/mode-new.md is the third place that states the cross-repos rule, and the test holds all three equal. A headless /plan new run under the budget is not yet measured, so #1922 stays open.

  • /evolve reads owner decisions (lead-session-orchestrator#948): evolve.extra-sources accepts kind: owner-decisions and turns a recurring deviation pattern from the export into an owner-preference candidate, only above the thresholds in skills/evolve/references/evolve-analyze-mode.md. The candidate passes the same confirmation as every learning, and such a learning never grants an approval.

  • Agent caps are enforced, not just written down (#1959): the new PreToolUse hook pre-agent-cap.mjs on Agent counts this session's running agents (subagent sidecars minus finished ones, plus same-message dispatches claimed under a lock), classifies them as reading or process-starting from the agent's tool list, and denies a dispatch above the cap from the accepted Auflagen file, else the Standard-Auflagen (2 + 2) once fallback_due_at has passed; without fleet integration there is no cap. Below 15 % free memory (released above 20 %) no new process-starting agent starts; unmeasurable memory says „nicht messbar“ and keeps the file cap. It only blocks starts, never kills, and fails open on its own errors.

    • The transcript is read only when the count can change the verdict (#1965): an engaged brake denies a process-starting dispatch before any read, so a slow parse under memory pressure can no longer run into the 5 s hook timeout (which lets the dispatch through), and the "below the cap" shortcut counts only the dispatch's own class. A dispatch from a subagent counts like the coordinator's: nested agents land in the same subagents/ dir (16 of 4734 subagent transcripts in 14 days started agents, every nested sidecar was there, measured 2026-11-06), so the cap counts the whole agent tree of a session. A nested agent is released by its terminal notification in the main transcript (44 of 45 in 14 days; the rest still waits for the 30-min TTL); this also frees its file scope in pre-task-scope-disjoint.
  • Context threshold for every session (#1960): post-tool-context-threshold.mjs (PostToolUse) reads the last main-chain assistant line of the session's own transcript and adds one hint at 700k („Übergabe-Datei vorbereiten, keine neue Welle mehr beginnen“) and one at 800k („am nächsten MR-Schnitt übergeben“). Session Config context-threshold: (enabled, prepare-k, handover-k; defaults true/700/800). Subagent tool calls never spend the hint; an unmeasurable context writes no marker. The fleet check-in accepts an optional kontext_k.

  • ux-evaluator rubric-v2: a person's judgment before the counters (#1961): six human checks per page and viewport (five-second, primary-action, dead-or-locked, empty-state, density, decoration) run before the mechanical checks, which stay only as a footnote and never as the overall verdict. A report without human-check verdicts for every page × {1440, 390} is incomplete (BLOCKED); the target is labelled live / production-build / dev-server. scripts/lib/test-runner/human-checks.mjs gives a deterministic floor from the AX snapshot (placeholder areas, dead or locked controls, competing actions). /test takes the rubric from the profile: web-gate uses v2, the mac profiles keep v1.

  • Issue balance per session (#1947): session-end prints „Issues dieser Session: +N angelegt, −M geschlossen, Saldo …“ with population and measurement time (scripts/issue-balance.mjs); anything unmeasured reads „nicht messbar“, never 0. LOW findings that get filed go into the session's one [Backlog-Sammel] collector; MEDIUM and above stay single issues.

  • „Entscheidungen für dich“ and „Nicht geprüft“ (#1950): mandatory sections (up to five points, „keine“ allowed) in the close report and the handover template; plan tasks may name an „Erwartung“ the report compares against.

  • The Stop notification is a per-user switch (#1936): owner.yaml notifications.stop: false (or env SO_STOP_NOTIFICATION=off) stops on-stop from sending the desktop notification (OSC 9 + OSC 777, "Session stopped — your turn"). The rest of the Stop path runs unchanged; before, the only way out was disabling the whole hook. Default stays on. on/off/1/0 are accepted as well as booleans, because YAML reads a bare off as a string. Measured 2026-11-03: with 10+ parallel sessions in Ghostty, which renders both sequences, every Stop raised two banners. Documented in docs/owner-config-schema.md and scripts/lib/owner-config.example.yaml.

  • Pre-push reuses a Full Gate receipt for the pushed tree (#1897, SO side of N4): before every route, .husky/pre-push runs run-quality-gate.mjs --receipt-check --pushed-sha <sha> in its temp clone, a read-only probe that passes the push on a green receipt for exactly this clean commit tree (exit 0) and otherwise exits 10 without running, emitting or writing anything, so the remote, partial and full routes run unchanged.

    • Only context: 'pushed-tree' records count: the hook's local fallback mints them with the new write-only --receipt-record (always runs the suite, never reads the store or navigator, writes only on a clean tree whose store is another checkout), the remote router after a verified green run (with the gate host's alias and Node). A Quality-wave or session-end receipt from the working copy never answers pre-push.
    • The key neutralises Node (gateDefinitionAnyNode), because the router already accepts any gate host and the hosts differ (measured 2026-11-03: m5 v26.13.0, win-wsl v24.21.0, M4 v24.22.1); the Quality wave and session-end keep the full definition. The store is the real repo's (--ledger-root); the partial gate and the mirror smoke never read or write receipts, and pre-push never touches navigator.
    • A hit carries reused_from: {tree, at, host, key: 'any-node'}, and every hook-run Full Gate event carries pushed_sha. Applies to main and tags too; CI stays authoritative. The hook feature-detects --receipt-record in the pushed tree's runner, so an older commit keeps the old invocation. Measured before: 195 of 200 Full Gate runs since 2026-10-20 were pre-push runs, 9 re-ran a tree already green within 12 h, 0 reused a receipt.
  • Trusted publishing (ADR-0018 route B) is built, opt-in (#1906): release.mjs --publish --trusted pushes the tag and the curated snapshot to GitHub, where .github/workflows/publish.yml publishes with provenance after the owner approves the run in the GitHub environment npm-publish; no npm token is involved. The workflow runs on a pushed tag v* only, and its guard job ends the run without publishing when the version is already on npm. scripts/lib/release-trusted.mjs verifies the receipt from public registry data (trusted-publisher marker, provenance workflow, ref, commit and tarball digest, per-file content comparison); --reconcile and --resume-tail use it for a trusted proof, and release --check --trusted reports the trusted-publisher evidence in the publish-route row (a plain --check shows only whether the workflow is armed). The receipt also requires the downloaded tarball's sha512 to equal dist.integrity; under mirror: curated, --resume-tail and --reconcile check the snapshot commit's Source-Commit: trailer before pushing the GitHub tag or confirming a receipt. Without approval in time the script ends awaiting-workflow (exit 1); approve, then --resume-tail. The default --publish stays route C. Pending: the owner's Trusted Publisher setup on npmjs.com, the first rehearsal and the later default switch.

  • Codex SessionEnd is wired (#1919): hooks/hooks-codex.json registers on-session-end.mjs for SessionEnd with the upstream maximum timeout: 3; SessionEnd is native and required in scripts/lib/codex/plugin-contract.mjs and check-hooks-symmetry; under SO_PLATFORM=codex the telemetry flush is capped at 700 ms so the hook stays under the 3 s cap. Measured live twice on codex-cli 0.160.1 (2026-11-02, plugin installed from the branch into a throwaway CODEX_HOME): the hook fired, released the lock on its own (the model turn ended in a 401, so Stop never ran) and left no registry entry; the second run raised no peer question (#1874 point 2). Interrupt stays unwired and unmeasured.

  • Codex version check (#1919): scripts/lib/codex/version.mjs (MIN_CODEX_VERSION 0.160.0, checkCodexVersion with the three states ok / warn / unmeasurable — a binary that cannot be run is reported as Codex-Version nicht messbar (…), never as a warning); codex-install.mjs imports it, and runtime-readiness --codex adds an advisory codex key and line without touching the exit code.

  • /plan research budget (#1922): every Explore agent of the research phase gets a stated budget (web agents at most 6 WebSearch and 6 WebFetch calls, 15 local calls, about 4 minutes; local agents the local figure), fixed in headless and autonomous runs; at the budget the agent stops and reports Not checked: …, which becomes the finding's Next check; the recap and the Epic comment carry a Research: line. Measured basis: one fresh-install /plan new run made 217 web calls in 40 minutes without reaching a PRD (2026-11-02, #1811).

  • One cross-repos resolution rule (#1922): /, ~/ and Windows drive paths are paths and used as written, any other entry is a bare name under ~/Projects/<name>, and a missing folder is a hint, never an abort — stated verbatim in docs/session-config-reference.md and session-start Phase 5 (pinned by tests/skills/cross-repos-rule.test.mjs), implemented by resolveCrossRepoEntry() in scripts/lib/config/cross-repo.mjs; the dispatcher (enumerate.mjs) resolves declared cross-repo.projects entries by the same rule under its confinement root and skips missing folders.

  • The Full Gate receipt also lives in the navigator gate-run ledger (#1897): on a clean tree (the content tree equals HEAD^{tree}), run-quality-gate.mjs --variant full-gate --receipt reads the navigator ledger as gate so:full-gate when the SO-local store has no receipt, and writes every real run there with quelle: so.

    • A green entry is written only when typecheck, test and lint all passed; a run with a skipped check writes nothing, so a ledger receipt always stands for three real runs. A reuse writes nothing either, because the ledger has no time window; SO checks the 12-hour window itself.
    • A reuse from the ledger is visible as reused_from: {tree, at, source: 'navigator', entry_sha256}. A dirty tree never reaches the ledger.
    • Without navigator (not installed, or SO_NAVIGATOR_BIN=off) nothing changes. Any other failure prints one stderr line and never changes the gate's result.
  • /eval names reused gate runs (#1897, L8): the verification-evidence and gate-health lines add how many passing gate events reused a receipt instead of running. The rubric-v3 formulas are unchanged.

  • release.mjs --publish stages the release and waits for the owner's approval (ADR-0018 route C, owner decision queue #7906, #1906): the checked archive goes through npm stage publish with the token from .env.local; the staged version is proven by the + pkg@version line or npm stage list, the publish proof carries staged: {at, stageId}, and the script stops with the next steps (owner approval on npmjs.com with 2FA, filed as an owner-handlung queue item; then --resume-tail --target X.Y.Z, which reconciles the registry before it tags or pushes). --publish --direct keeps the pre-2027 route. The npm-token-publish row no longer demands Bypass-2FA for the staged route (stage-capable), and the publish-route row has no January-2027 sunset for it; both stay as before under --direct. A new preflight row staged-proof-absent refuses a second --publish while the proof carries a staged block; an exit 0 of npm stage publish that matches neither receipt signal still writes the proof with staged.confirmed: false before it reports, so --resume-tail can finish after the approval. ADR-0018 is Accepted: route B (Trusted Publishing from the curated mirror) is due by 2026-12-15.

  • Behaviour test for the site's command rewrite (#1879): tests/site/commands-rewrite.test.mjs runs the real site/assets/commands.js inside node:vm with a stub document and checks the MENTION rewrite through an inert window.__soTest seam; removing the lookahead boundary, the lead group or the ,/) members turns it red; no DOM library added, the script's cache-bust bumped to ?v=20261102.

  • The owner-config probe names a loose owner.yaml (#1636 point 8): when the file carries any group/world bit or its directory is group/world-readable or -writable, the session-start banner prints the measured modes and the exact chmod 700 '<dir>'; chmod 600 '<file>' fix (paths shell-quoted); silent when tight, on win32, when the file is absent or on a stat error; the probe's telemetry carries loose_mode: 0/1.

  • Three security-delta contract tests (#1912): an already-expired deadline sends no OSV request and reports degraded/timeout with the unchecked count; a detail-fetch timeout is classified deadline and a connection reset network, the finding kept with severity unknown; the CLI text line prints unchecked=<n> for a degraded check. The remaining acceptance bullets of #1912 were already covered (182 tests, measured 2026-11-02).

  • Codex apply_patch writes reach the Edit/Write guards (#1913): enforce-scope, config-protection, the sessions-ledger guard and the STATE.md guard now run on every file a Codex apply_patch touches, through the fail-closed adapter hooks/codex-apply-patch-guard.mjs (patch parsing in scripts/lib/codex/apply-patch-parse.mjs). It is wired as the apply_patch PreToolUse group in hooks/hooks-codex.json and listed in SAFETY_HOOKS. The payload shape was derived from the codex-cli 0.160.1 binary and an external report, not from a live-logged payload; logging one is the follow-up #1914. Proposal and report by GitHub user dviolante (GH#83); the implementation was written here.

  • ADR-0017 fixes the destructive guard's target and non-goals (#1854, owner decision queue #5975 option 1): the guard protects against irreversible loss of uncommitted or untracked work and shared working-copy state (the rule ids in .orchestrator/policy/blocked-commands.json); lexical variants the tokenizer can read and frequently measured forms are in goal, payloads it cannot see (variable payloads, other interpreters, xargs, find -exec, script files, nesting past the caps) are non-goals. The open guard follow-ups are classified against it.

  • check-direct-rename validator (#1865): validate-plugin fails when a file under scripts/ or hooks/ gains a direct rename/renameSync call outside the shared atomic writers; a per-file allowlist (24 calls in 19 files) can only shrink, and an intentional non-write rename carries // direct-rename:ok <reason> on the same line or a comment-only line above.

  • so-hud outwardHeadless: deny (#1817): an opt-in userConfig switch (default allow) that refuses outward actions (push to main, force, tags, merge, release, publish, deploy) in runs with no surface instead of letting them through with a receipt; the refusal writes receipt ohne-zeichenflaeche-abgelehnt and names the way to allow the call. A hurdle, not a guarantee: protected branches and CI stay authoritative. Claude Code only.

  • release --check row publish-route (#1906): reports which publish route is configured (token, trusted-publishing, none); the token route turns red from 2027-01-01, when npm ends direct publishing with Bypass-2FA tokens, with an injected clock (SO_RELEASE_NOW) for rehearsals. The OIDC workflow .github/workflows/publish.yml (tag guard, npm audit signatures) was added as an inert draft here and is armed on tag push by the route B entry above; the owner decision is queue #7873.

  • .claude/rules/bash-harness-pitfalls.md § 7 (#1890): writer | grep -q under pipefail is randomly red (SIGPIPE 141); use grep -q PATTERN <<< "$var" or a file.