voicePace and readingPace set each row's floor, span and duration in one place: the spoken rest reveals the unread words over the speech audio, a procedure word reveals none, a burst and silent reading reveal the rest at reading pace, and wordsRead never goes below the floor. A sound that fails or is dropped still ends cleared. Requirement 003 W6 states it; probes cover quiet, burst, fallback, cut, auto, check, silent, rejected and faulty rows.
🎯 Fix: 92 %
💥 Break: 4 %
💸 Tokens: 100 %
⏱️ Velocity: 100 %
🛡️ Security: 100 %
🏆 Score: 88
🚦 Gate: ✅ 90/5, tokens and velocity <= 100, security 100 %
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
rasci.py artifacts takes --worker <text> to list only that worker's items. set refuses an --artifact or --drop-artifact without <kind>:<ref> (it used to no-op silently on a bare path), and a well-formed drop that is not recorded drops nothing and says so on stderr, so idempotent teardown runs keep working.
🎯 Fix: 93 %
💥 Break: 3 %
💸 Tokens: 100 %
⏱️ Velocity: 100 %
🛡️ Security: 100 %
🏆 Score: 90
🚦 Gate: ✅ 90/5, tokens and velocity <= 100, security 100 %
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
hotfix.py add now adopts an existing rasci task of the same branch and repository: the adding session becomes R and A, the former A keeps its other letters, and a task on another branch or repository is refused instead of overwritten.
🎯 Fix: 92 %
💥 Break: 4 %
💸 Tokens: 100 %
⏱️ Velocity: 100 %
🛡️ Security: 100 %
🏆 Score: 88
🚦 Gate: ✅ 90/5, tokens and velocity <= 100, security 100 %
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
score.py and backfill.py ran with python -I drop their own directory from sys.path, so the sibling imports failed (Security 2.0 hit it 2026-11-14 02:40). Both now put their resolved script directory first; tests run both scripts isolated.
🎯 Fix: 95 %
💥 Break: 2 %
💸 Tokens: 100 %
⏱️ Velocity: 100 %
🛡️ Security: 100 %
🏆 Score: 93
🚦 Gate: ✅ 90/5, tokens and velocity <= 100, security 100 %
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The popup probe clicked elementFromPoint(2, innerHeight - 2) to close the popup. In Firefox at 1920 px that corner returned the html element in 4 of 10 stress rounds, so the close handler, which acts only on a click on the dialog itself, correctly ignored it. The probe now clicks the middle of the widest free margin around the popup box and waits up to 3 s until that point hits the dialog, else clicks whatever is there, so a real overlay still fails.
🎯 Fix: 85 %
💥 Break: 5 %
💸 Cost: –
⏱️ Time: –
🛡️ Security: 100 %
🏆 Score: 81
🚦 Gate: ⛔ fix 85 % < 90 (operator word 2026-11-13: commit and merge)
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
At 390 px with the radio spot open on Agents Orbit, the STARS probe reported the legend star covered by the radio search field and its clear button. Both sit scrolled out of view inside the radio spot's overflow box, and a hit test at their points returns the legend, so nothing covers the star. The probe now intersects each box with every clipping ancestor before it measures overlap; it can only shrink a box, so a visible overlap is still reported.
🎯 Fix: 93 %
💥 Break: 4 %
💸 Cost: –
⏱️ Time: –
🛡️ Security: 100 %
🏆 Score: 89
🚦 Gate: ✅ 90/5, cost and time <= 100, security 100 %
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The 008 star row took about 26 px from the stage cards: the tasks/3d canvas fell under 80 px, where zoomToFit's 40 px padding gave a negative field of view and froze the camera, and the agents/orbit card pushed the page past 844 px with the radio spot open, against 001's no-scrollbar rule. The chart legend is now the panel's star slot (style-home.css places it at the end of the last key line under 640 px), and the 3D fit clamps its padding to a quarter of the canvas height and skips a canvas of height 0. The Firefox radio probe waits for paused with src cleared, since Firefox pauses the element before the product's 3 s fade ends.
🎯 Fix: 92 %
💥 Break: 4 %
💸 Cost: –
⏱️ Time: –
🛡️ Security: 100 %
🏆 Score: 88
🚦 Gate: ✅ 90/5, cost and time <= 100, security 100 %
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Requirement 008 put Home first in nav.js, so SCREENS and the walk badges gain home, key 1 now opens #home and Resources moves to key 6 (page-keyboard.js numbers keys by tab order). TestHomeDashboard borrowed TestOrchestratorE2E.setUp, whose zero-argument super() raised TypeError, so it never ran; the fixture is now TestOrchestratorE2E.seed(), called by both. Its first real run needed the two-argument speech stub, the gone-tile step at the last width, an overlap check that skips closed details menus and popup layers, and the grid expectation style-home.css really computes.
🎯 Fix: 93 %
💥 Break: 3 %
💸 Cost: –
⏱️ Time: –
🛡️ Security: 100 %
🏆 Score: 90
🚦 Gate: ✅ 90/5, cost and time <= 100, security 100 %
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Requirement 004 (acf3925) moved speech on, classes and muted into each browser's localStorage and gave the speech tool a second argument admit; the bubble-faces fixture still saved the old hub settings and stubbed the one-argument tool, which settings.py now refuses. The fixture saves only bubbles, stubs (body, admit) and sets the radio to Silence with mayday and security through a start-up script. W6's long row moves from security to broadcast, the class 004 switched W6 to, so the browser picks it again.
🎯 Fix: 95 %
💥 Break: 2 %
💸 Cost: –
⏱️ Time: –
🛡️ Security: 100 %
🏆 Score: 93
🚦 Gate: ✅ 90/5, cost and time <= 100, security 100 %
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The radio_module test wrote five fake radio.py versions with nanosecond mtimes, all in second 0. Versions 5 and 6 have the same size, so with bytecode writing on (make test) CPython's pyc check reused version 5 for version 6 and the test failed only in the full suite. The installed hub is not affected: install.sh drops every pyc, the container sets PYTHONDONTWRITEBYTECODE and mounts /skills read-only.
🎯 Fix: 95 %
💥 Break: 1 %
💸 Cost: –
⏱️ Time: –
🛡️ Security: 100 %
🏆 Score: 94
🚦 Gate: ✅ 90/5, cost and time <= 100, security 100 %
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
937c2c0 added the line prefix KEPT to watch.py, which shadowed the state directory path watch.KEPT that gates.py and the queue cleanup use; the prefix is now KEPT_LINE. The receipts page test gets an identity revealed() since page-dom.js wraps names through it after the firewall merge, the operator tasks test expects the Home tab of 008, and the hub view id fixture uses a non-ASCII word the English check accepts. Accepted MEDUSA false positives, re-keyed by the rename: MM-SCAN-058 at watch.py:1417 and WM-SCAN-354 at watch.py:1418.
🎯 Fix: 92 %
💥 Break: 2 %
💸 Cost: –
⏱️ Time: –
🛡️ Security: 100 %
🏆 Score: 90
🚦 Gate: ✅ 90/5, cost and time <= 100, security 100 %
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The browser suites now check the tab and panel labels that requirements 001 to 003 list but no e2e test reached: every hub tab hash including #home and #resources, and the workbench sheets without an accessible name. Committed below the bar on the operator's word; the unify e2e run proves it.
🎯 Fix: 75 %
💥 Break: 8 %
💸 Cost: –
⏱️ Time: –
🛡️ Security: 100 %
🏆 Score: 69
🚦 Gate: ⛔ fix 75 % < 90, break 8 % > 5
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
passkey.py accepted an RS256 public key of any modulus size. It now refuses a modulus under 2048 bits at registration and at every verification, before either signature backend, with a fixed error asking for a new passkey; the library backend checks key_size again and the openssl backend's key_type requires the same size. The SubjectPublicKeyInfo is read as strict DER (no negative or padded INTEGERs, minimal lengths, no unused bits, no trailing data) and the rsaEncryption identifier must match exactly. ES256 and EdDSA are unchanged. An already pinned weak key is refused on every use until a new passkey is pinned. Accepted MEDUSA findings: AIC-EXFIL-005 and PRIV-SCAN-762 at tests/hub_support.py:151 (public half of a throwaway test key) and PQC001 at passkey.py:236 (WebAuthn offers no post-quantum algorithm yet).
🎯 Fix: 95 %
💥 Break: 3 %
💸 Cost: –
⏱️ Time: –
🛡️ Security: 100 %
🏆 Score: 92
🚦 Gate: ✅ 90/5, cost and time <= 100, security 100 %
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Requirement 001 S6: the hub gets a Firewall view of the Claude Code permission and sandbox settings. The operator signs each change with the passkey; a host settings writer (systemd path unit) applies only exact, signed changes to allowlisted settings files through O_NOFOLLOW directory and file descriptors, and a timer refreshes the view. Every signed text and every page sink refuses or marks hidden characters: Cc, Cf, Zl, Zp, every space but U+0020 and the Default_Ignorable set, with U+FE0E/U+FE0F allowed only after one of the 359 non-ASCII emoji bases of Unicode 16 (shared table, validated at import, the same pair rule on the page). Removal of a deny, ask or denyRead entry is exempt only as an exact single remove. The permission hook prints the hub's refusal reason, cleaned of hidden characters. Accepted MEDUSA false positives (independent review, operator pick): 38 this round, mainly the Unicode defence tables read as attack patterns, generic MM-SCAN-058 heuristics, WEB-FLASK-004 at view.py:213 (ints from the validated table via tojson), EXF26-TOOL-004 on the non-secret WebAuthn challenge, PRIV-SCAN-060 on the office-action shape check and PQC007 on a picker key.
🎯 Fix: 95 %
💥 Break: 3 %
💸 Cost: –
⏱️ Time: –
🛡️ Security: 100 %
🏆 Score: 92
🚦 Gate: ✅ 90/5, cost and time <= 100, security 100 %
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Requirement 005: every radio message in the Stream and the radio log shows the avatars of the agents that read it, with read time and an ack badge where a receipt exists; receipt rows no longer appear as entries of their own, and the sender shows once. Read times come from seen marks, stamped only after a fresh presence pass no older than two sighting intervals; a stored time before its row is refused. An operator message links to its relay row through the ts radio.py prints first on stdout right after store(), checked against an exact row (sender Operator, relay text, targets, no to_sessions); a ts shared by two rows shows no readers. radio.py clips printed addresses; the hub keeps the first stdout line whole. Accepted MEDUSA false positives: MM-SCAN-058 at server.py:73, JB-SCAN-641 and JB-SCAN-616 at radio.py:389 (clip, the output sanitizer), FLA-SCAN-055 at radio.py:109. Committed below the 90/5 bar on the operator's word; the receipts e2e runs through the queue.
🎯 Fix: 89 %
💥 Break: 6 %
💸 Cost: –
⏱️ Time: –
🛡️ Security: 100 %
🏆 Score: 84
🚦 Gate: ⛔ fix 89 % < 90, break 6 % > 5
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Resolution: rules, limits, line/file counts and not-covered judge the aliased tree path medusa scans and name the repository path; findings, cache key (tree path, repository path, not covered) and accept rows keep the real path.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Requirement 001 S8: the hub can use a shared OpenAI-style speech service (KVB_ORCHESTRATOR_SPEECH_URL, key from a mounted file) instead of the local piper and whisper engines, and the image takes its pinned artefacts through an optional INFINITO_ARTEFACT_BASE cache. A new transport serves speech and the LLM: one wall-clock deadline over DNS, connect, TLS handshake and body, certificate and hostname always verified, no redirect followed. A key travels only over https or to loopback; both the speech and the LLM start refuse it otherwise. Key files are capped at 4096 bytes. Streamed WAV answers with placeholder sizes are accepted only when the data chunk runs to the end, and are rewritten before caching. Accepted MEDUSA false positives (41, independent review, operator pick): raw client socket patterns in transport.py, AIC-EXFIL-005 on the sealed key path, JB-SCAN-2237 on the hub's own system prompt, STG009 on the RIFF pad byte and LLM01 at language.py:111, where agent-written titles reach the NL prompt but the model only proposes a plan the operator confirms.
🎯 Fix: 93 %
💥 Break: 4 %
💸 Cost: –
⏱️ Time: –
🛡️ Security: 100 %
🏆 Score: 89
🚦 Gate: ✅ 90/5, cost and time <= 100, security 100 %
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Operator order 2026-11-13: once the operator ran a command, the agent that queued it must hear it directly, without the operator relaying it. Every row close now sends one radio call to the stored initiator (party name plus the session id add pinned; legacy rows by the A name): run and hub ran (done, exit 0), check still open, the check-pass sweep of list and run, done, cancel, failure and hub failed. The call names row, outcome and exit code and never carries command output; shared stderr stays in failed.output of the register. radio.py send takes --sid for a call to exactly one address. Accepted MEDUSA false positives: CGS-SCAN-029 and CGS-SCAN-030 at operator.py:930, which runs the row command only after the operator's per-row answer.
🎯 Fix: 93 %
💥 Break: 4 %
💸 Cost: 105 %
⏱️ Time: 100 %
🛡️ Security: 100 %
🏆 Score: 85
🚦 Gate: ⚠️ cost 105 % > 100: ask the operator before the commit
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
security.py cached any run that exited cleanly, so a run MEDUSA cut short by its file size, line or path limits, or one that skipped an unscannable type, was stored and reported as 100 %. It now predicts which changed files MEDUSA takes, from MEDUSA's own limits and exclusion defaults read from its source, and caches a result only when MEDUSA reports exactly that count. Files it does not scan are listed as not covered; a diff with nothing scannable, an unreadable limit or a count mismatch exits 4 instead of scoring 100. Coverage details move to coverage.md, keeping SKILL.md under the word cap.
🎯 Fix: 95 %
💥 Break: 2 %
💸 Cost: 93 %
⏱️ Time: 100 %
🛡️ Security: 100 %
🏆 Score: 100
🚦 Gate: ✅ 90/5, cost and time <= 100, security 100 %
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Requirement 006 grew SKILL.md to 853 words, over the 600-word cap test_lock enforces. The tables and message block details move to tables.md and the score journal with its backfill to journal.md; SKILL.md keeps the commit-block rule and points to both. The journal reader set in test_score_journal follows the text to journal.md.
🎯 Fix: 95 %
💥 Break: 2 %
💸 Cost: 110 %
⏱️ Time: –
🛡️ Security: ~100 %
🏆 Score: 85
🚦 Gate: ⚠️ cost 110 % > 100: ask the operator before the commit
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The stop-stage cleanup ran the configured docker image prune, which also removed dangling images still pinned by digest, so the next deploy pulled them again. A guard shim now sits first on PATH for every configured cleanup line: a plain docker image prune (also -a) becomes one pass that removes only images whose repository and digest are both <none>, without force, on the daemon the call names; an image a container uses is kept. system prune, rmi, image rm and compose down --rmi are refused and reported. Worktree cleanup commands run as argv lists; their nested image prune gets the same leftover pass inside the worktree daemon. The guard protects operator-written lines against accidents, not against a script that bypasses PATH.
🎯 Fix: 95 %
💥 Break: 3 %
💸 Cost: 106 %
⏱️ Time: –
🛡️ Security: 100 %
🏆 Score: 87
🚦 Gate: ⚠️ cost 106 % > 100: ask the operator before the commit
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
006's label test scanned docs/requirements too, but the 006 requirement has to name Cost and Time to describe their rename, so the test failed once the requirement and the code met on the unify branch. The test now skips docs/requirements, and the 006 acceptance line says so.
🎯 Fix: 96 %
💥 Break: 1 %
💸 Cost: 100 %
⏱️ Time: 100 %
🛡️ Security: 100 %
🏆 Score: 95
🚦 Gate: ✅ 90/5, cost and time <= 100, security 100 %
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A Home tab left of Stream shows every panel and item the operator starred: a star on every panel and item pins it, tiles sit on a 12-column grid, resize in width and height, reorder by drag, render live, and a pinned item that no longer exists shows a gone tile with an unpin button. Home becomes the start tab once something is pinned. The layout lives in 007's view store, changed only through update() so two tabs never overwrite each other's pins. Committed below the 90/5 bar on the operator's word so the unify e2e proves it before master moves.
🎯 Fix: 88 %
💥 Break: 6 %
💸 Cost: 100 %
⏱️ Time: 100 %
🛡️ Security: 100 %
🏆 Score: 83
🚦 Gate: ⛔ fix 88 % < 90, break 6 % > 5
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The Resources tab draws one hardware chart and one tokens chart with an accumulated switch, plus a combined overlay of hardware, tokens and the burnup, every series switched by a legend chip and all shown by default. The hub keeps every toggle in a validated view store: PATCH /api/view takes a set and the expected old values and answers 409 on a stale expectation, update() re-runs a change on the stored value, list paths change only through update(), and a page shown again from the back-forward cache fetches the stored view anew.
🎯 Fix: 93 %
💥 Break: 4 %
💸 Cost: 100 %
⏱️ Time: 100 %
🛡️ Security: 100 %
🏆 Score: 89
🚦 Gate: ✅ 90/5, cost and time <= 100, security 100 %
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
score.py names the Cost and Time metrics Tokens and Velocity from one label file and appends every scored block to a score journal in the council root. backfill.py rebuilds that journal from commit messages with its own label mapping, deterministic and idempotent, so another host can import its history. The hub serves the journal, and the Gantt burndown draws each metric as its own overlay on a right 0-100 % axis, every overlay off by default and switched by legend chips whose state the view store keeps.
🎯 Fix: 90 %
💥 Break: 5 %
💸 Cost: 100 %
⏱️ Time: 100 %
🛡️ Security: 100 %
🏆 Score: 86
🚦 Gate: ✅ 90/5, cost and time <= 100, security 100 %
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
README.md conflict with the queue-cleanup timer sentence resolved by keeping both: the unify paragraph plus the council deny sentence. install.sh auto-merged (queue-cleanup state dir, pycache prune, council-deny call), shellcheck clean.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
New scripts/council-deny.js holds the deny rules (SPOT) for ~/.claude/agents-council/** and, new, ~/.local/state/agents-council/**, where the skills' scripts keep state only they may change. enable-plugins.js takes them from it for <TARGET>/.claude/settings.local.json as before. The global install now also merges them into ~/.claude/settings.json, the only user-wide settings file, because a settings.local.json binds only sessions rooted in its own repository; the operator no longer edits settings by hand. The writer follows a settings symlink (dangling included), keeps the file mode, writes a 0600 wx temp file and renames it, and refuses invalid JSON or a non-object with exit 1 and the file untouched.
Tests: tests/test_dream.py 26 passed (denies once beside existing rules, user settings kept through a symlink with mode, dangling link kept, invalid JSON untouched with exit 1); two independent reviews, round 2 holds.
🎯 Fix: 93 %
💥 Break: 3 %
💸 Cost: 100 %
⏱️ Time: 100 %
🛡️ Security: 100 %
🏆 Score: 90
🚦 Gate: ✅ 90/5, cost and time <= 100, security 100 %
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
install.sh copies the rendered skills with cp -a, which carried __pycache__, *.pyc and *.pyo along; Python loads a lone .pyc before the module of the same name, so a stray file in a skill could shadow its script. After each copy, find -H prunes them in .agents/skills and .claude/skills, also when .claude/skills is a symlink to .agents/skills; the delete before the copy uses -H too. orchestrator.sh drops .pyc and .pyo besides __pycache__ in the hub tree.
Tests: test_no_python_bytecode_reaches_the_installed_skills (linked and unlinked) and test_install_ships_no_bytecode_from_the_tree fail on base 3418738 and pass; shellcheck and ruff clean; two independent reviews, round 2 holds.
🎯 Fix: 95 %
💥 Break: 2 %
💸 Cost: 100 %
⏱️ Time: 100 %
🛡️ Security: 100 %
🏆 Score: 93
🚦 Gate: ✅ 90/5, cost and time <= 100, security 100 %
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
radio.py conflict resolved by the unify side plus the prepared resolution (seen() on top of the feedback-channel radio.py); receipts.py and its tests adapted to the unify helpers. Merged tree: test_hub_receipts 15, test_radio_seen 8, test_radio 45 passed.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
receipts.py builds, for every radio row in a window, the sessions that took it (from their prompt and radio-seen marks) and those that acknowledged a call, with names taken only from the host agent-sessions.json snapshot, so a forged name in a row never reaches the readers; one snapshot build per call keeps a 90k-entry history at 13 ms instead of 8385 ms. radio.py gains seen(), which pending() now uses, so the hooks and the hub decide taken the same way. gating.py and receipts.py share one script loader (skill_scripts.script_module). Not wired into the hub yet: phase B.
Tests: tests/test_hub_receipts.py and tests/test_radio_seen.py; both import unify-side helpers and run green merged into unify/skills-20261112-night with the prepared merge resolution.
🎯 Fix: 92 %
💥 Break: 3 %
💸 Cost: 100 %
⏱️ Time: 100 %
🛡️ Security: 100 %
🏆 Score: 89
🚦 Gate: ✅ 90/5, cost and time <= 100, security 100 %
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The hub gates uncached /api/tts renders on the latest W4 host CPU sample (503 busy or unknown, cached lines always served, signed speech.threshold 30 to 90 in steps of 10, speech.gate in /api/state). The page plays a WebAudio squelch-tail burst when no speech plays, offers Auto, Quiet and Silence per browser, and always sounds the MAYDAY, PAN-PAN and SECURITE procedure word. The radio panel becomes a minimalist VHF faceplate with display line, mode buttons, knobs, PTT send and the internet station; the settings dialog loses its radio section. The log gets Sent, Received and All tabs with class, sender, recipient and agent chips plus text search. Decisions 3 and 10 of 004 now state that the procedure word still sounds in Silence.
Security: MEDUSA-MEX-SCAN-1167 at orchestrator/kvb_orchestrator/hub.py:126 accepted as a false positive after the round 5 skeptic review and the operator pick of 2026-11-13 (regex hit on rce inside RESOURCES; CpuGate reads host CPU samples, no model).
Break risk 6 % is above the 5 % bar; committed and merged on the operator pick of 2026-11-13 for this bundle.
🎯 Fix: 92 %
💥 Break: 6 %
💸 Cost: 100 %
⏱️ Time: 100 %
🛡️ Security: 100 %
🏆 Score: 86
🚦 Gate: ⛔ break 6 % > 5
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>