## (daily refresh, 1)

- cross-category: stars scan clean, 3 new stars since checkpoint (tudorms/BookScan, bookorbit/bookorbit, pnpm/pnpm), all rejected out of scope; checkpoint holds at 2026-11-04T07:47:57Z.
- cross-category: andyrewlee/awesome-agent-orchestrators and Agent-Analytics/awesome-multi-agent-orchestrators churned (2026-11-04 and 2026-11-05 commits), full walks done; ai-for-developers, hesreallyhim, e2b-dev, slavakurilyak unchanged (hesreallyhim content stopped 2026-10-22, rest predate first checks), scans carried over.
- cross-category: open-swe (langchain-ai, 11.2k stars, MIT) flagged for orchestration (or executions), strongest new entrant.
- cross-category: concord-mcp (952 stars, MIT) flagged for orchestration per Foremerge precedent, protocols alternate.
- cross-category: Cotal (381 stars, Apache-2.0) flagged for protocols.
- cross-category: zclaw (2,241 stars, MIT, dormant since 2026-05-17) flagged for assistant-runtimes as smallest compile target (888 KiB ESP32).
- cross-category: speakeasy gram (277 stars, AGPL-3.0) flagged for control-planes (IBM ContextForge analog).
- cross-category: centaur (paradigmxyz, 1,421 stars, license unverified) flagged for orchestration.
- cross-category: aiwg (jmagly, 228 stars, MIT, owner-starred) flagged for skills.
- cross-category: AMP (agentmessaging/protocol, 37 stars, Apache-2.0) flagged for protocols per ANP/PIC thin-spec precedent.
- cross-category: OrcaReplay (340 stars, Apache-2.0, quiet since 2026-10-01) flagged for session-analytics.
- cross-category: agent-barn (111 stars, Apache-2.0) flagged for control-planes.
- cross-category: vendor-sibling folds proposed, berd to goose, okto-nexus to Okto Pulse, claudexor to Ouroboros, lettabot succession to Letta.
- cross-category: notch-first surface Perch rejected under the mission-control-tails ruling; therustyclaw repo Not Found as of 2026-11-05; OpenHuman stays on the absorbed-family shelf.
- cross-category: about 140 below-bar rejections logged (me-too ADE bulk, content-class packs, dormant -claw verticals, thin-traction candidates).
- section-root: model-selection-for-coding-tasks: Kimi K3 OpenRouter readings split their floors across hosts (the aggregator listing's input leg re-priced back down from $0.20 to $0.192/$16.00 with cache reads at $0.182, Wafer's fast row re-priced down to $0.153/$22.00 and reclaiming the input floor, Relace's fp4 row re-priced up to $0.1539/$10.5539 and losing both legs, the output floor falling to Chutes' mxfp4 row at $1.85/$9.50, no single row holding both legs, the roster widening to 31) while kimi-k2.7-code's floor held at $0.60/$3.00 on CoreWeave's int4 row.
- section-root: model-selection-for-coding-tasks: OpenAI's page extended Sol's promo end from 2026-11-21 to 2026-11-30 with gpt-5.6-terra and gpt-5.6-luna still listed at launch prices, and DeepSeek's page replaced its free-through-October Search-and-Fetch wording with a standing per-request section at $0.75 per 1,000 requests each, five days past the stated 2026-11-01 start.
- section-root: model-selection-for-coding-tasks: Kimi K3.1 Pro's full open weights confirmed shipped (Moonshot's quickstart presents the model as open-weight with the pending-weights caveat gone, models.dev marks it Open, and the moonshotai/Kimi-K3.1-Pro Hugging Face repo serves the release).
- section-root: model-selection-for-coding-tasks: Google, Anthropic, Z.ai, and Moonshot pricing pages re-fetched with every cell unchanged, the Aliyun billing page again serving its Chinese locale with the batch and cache terms intact and the USD rates corroborated by OpenRouter and models.dev, which showed no new covered-provider releases after the 2026-10-26 terra rung.
- section-root: model-provider-feature-matrix: Moonshot flagship and workhorse cells synced to the day's OpenRouter readings, the OpenAI windows cell moved to Sol's promo extension, and the DeepSeek windows cell moved to the standing Search-and-Fetch pricing section; every other vendor cell re-verified unchanged.
- section-root: agentic-coding-tools-landscape: DeepSeek Harness's newest prerelease moved to dsh-v0.2.2-rc.1 (published 2026-11-04, the 0.2.2 train's first release candidate).
- section-root: agentic-coding-tools-landscape: every cited external link re-fetched, sixteen of eighteen fetched 200 with Bloomberg and devops.com again 403 to scripted fetches; the Gemini CLI enterprise-docs stamp held at 2026-10-26 with the same banner and the Cursor pricing page re-read with the X-bundle plans unchanged.
- section-root: the-agentic-development-environment-landscape: Paradev relicensed from Elastic License 2.0 to MIT and reworded its README pitch, moving it into the open-source group on the leaderboard paragraph, the top-five mix line, and the license axis.
- section-root: the-agentic-development-environment-landscape: Paseo's Apache-2.0 claim re-verified (its custom LICENSE wrapper names Apache-2.0 for the project's own code, which is why GitHub's API reports NOASSERTION), Agent Orchestrator's supported-agent table re-counted at 36 so the 35-plus claim holds, Air's supported-agent, surface, and cloud-run claims re-verified unchanged, and all 17 URLs fetched 200.
- section-root: model-benchmark-matrix: every cited URL re-fetched, all fetched 200 except the OpenAI PaperBench page again 403 to the scripted fetch, and Terminal-Bench still at 4.0.
- section-root: the-tells-are-structural: all eight cited sources re-fetched 200 with the study figures unchanged and the no-ai-slop ruleset's structural rule ids intact.
- section-root: context-management-patterns: all nine cited pages fetched 200 with the compaction-survival, skill-cap, idle-compaction, v2.1.304, and /autocompact claims holding on the live context-window page.
- arbor: recorded the November 3 batch (compaction tool-exchange preservation, empty-reply skip, Serply Google Scholar backend) with the release train still paused at v0.1.4
- aris: recorded the November 1-3 docs pushes (Absent Author News entry, Goal Pulse activation route, HERO handoff connection)
- deepanalyze: recorded the November 2 README overhaul (DeepAnalyze.net and sponsors promotion, video fixes)
- kapso: recorded the v0.8.0 breaking redesign (verbs renamed, modes replaced by one configuration, Python API rebuilt around kapso.build()), rewrote the pillar descriptions to the renamed surfaces, harmonized the cautions' star figure, and noted main's November 3-4 resume fixes
- local-deep-research: recorded the November 2-5 security and cache hardening batch, release train still at v1.10.7
- mole: recorded the November 1 doctor test-hygiene PR and reworded the README weak-spots table reference (banned-term cleanup)
- openresearch: recorded v0.2.24 (November 4) and main's November 4-5 experiment-engine rewrite and subagent rework
- autocontext: recorded the November 3-5 CPU milestone wave (checkpoint editing, weight reproduction, measured outcomes), refreshed commit counts, and re-confirmed the PyPI pin gap
- agon: rotation re-verification, all references current; folded the AutoResearch-Factory siblings AutoSR and AgonAlpha as ecosystem evidence
- coral: rotation re-verification, all references current, no movement since October 19 on dev and September 6 on main
- gpt-researcher: rotation re-verification, all references current, v3.7.0 still the latest release
- data-to-paper: rotation re-verification, all references current; reworded the self-grading sentence (banned-term cleanup)
- autosci: reworded the internal-beta sentence (banned-term cleanup); no repository movement since October 30
- git-lrc: pushed_at moved to November 3 on a side branch, main stays October 23, v0.6.17 still latest
- kodus: recorded the November 2 fix batch (2.2.10, web-1.2.11, selfhosted-2.2.8) and the November 3 batch (2.2.11, web-1.2.12, selfhosted-2.2.9, mcp-manager-2.1.6: per-endpoint BYOK keys, Azure Repos Kody Rules evaluation, MCP proactive hints in tool _meta), push November 5
- multi: recorded the November 2 v1.21.0-1.21.2 train (omp and OpenCode as native plugin hosts, harness detection fix, agy version-manager fix), 31 tags, v1.21.2 latest
- open-code-review: recorded v1.13.4 (November 2) and v1.13.5 (November 3, HAML/Slim/ERB/EJS/Liquid/Nunjucks templates, jsonc, Scala rules), ledger at 143 releases, push November 5
- openqodex: push moved to November 5 on side branches (main stays November 1), npm backfill raised the October 25-31 week to 1,471, trailing month 8,244 to November 5
- qodo: push moved to November 5, v0.49.0 still latest
- qodo: See-also matrix link corrected to the Code Review Feature Matrix
- roborev: push moved to November 5, v0.74.0 still latest
- sourcery: rotation re-verified, corrected repo liveness (main quiet since September 17, later pushed dates side branches, 1.46.1 betas unreleased on PyPI), customer strip grew to eight logos (Red Hat, Linux Foundation), prices and PyPI unchanged
- graphite-diamond: rotation re-verified, vendor launch and joins-Cursor posts removed from graphite.com (404), archived joins-Cursor copy cited, prices and live surfaces unchanged
- coderabbit: See-also matrix link corrected to the Code Review Feature Matrix
- code-review: matrix refreshed (OpenCodeReview 143 releases, multi 31 tags, OpenQodex npm 8,244 trailing month), links, front matter, and as-of dates verified, category fit unchanged
- code-review: kodustech/awesome-ai-code-review still unwired as a mapped source (pushed_at still 2026-06-28), q21 evidence updated
- context7: recorded the 2026-11-03 releases (mcp 4.5.0 shipping the research-preview mode with its submit-docs-feedback tool, ctx7 0.9.0 adding cross-library `ctx7 search` and OpenCode 2.x config handling), trailing-month npm 2,769,609 to 2026-11-05, pricing unchanged
- entroly: pushed signal resolved non-material (unreleased 2026-11-02 proxy fix, release still 1.0.88), note untouched
- graft: recorded main's unreleased 2026-11-02 fixes (peak memory on big repos, a sync crash), npm unchanged at 0.28.1, counts to 2026-11-05
- graphify: recorded releases v0.9.100 to v0.9.102 (2026-11-01 to 11-03: node-id collision fix, C++/Rust/Python/Go resolution accuracy, query-layer and incremental-update correctness), 2,831 commits, 369 contributors
- headroom: pushed signal resolved non-material (2026-11-05 fix sweep unreleased, PyPI still 0.42.0), note untouched
- jevgrep: full rotation re-verification, all citations intact, no release past v0.8.5, trailing-month npm 22,007 to 2026-11-05
- repomix: pushed signal resolved non-material (test-infrastructure and skill-fix commits unreleased, v1.18.1 latest), note untouched
- repowise: full rotation re-verification; recorded the 0.60 train (v0.60.0 2026-11-01: plugin API with author guide and PLUGIN_API_VERSION; v0.60.1 2026-11-02: decision folding, confirm-a-decision-with-reason, update phase timings)
- rtk: dev-1.2.0 candidate train at rc.615 as of 2026-11-05, stable remains v1.1.0
- semble: recorded v0.7.2 (2026-11-03, generated skill description made valid YAML), PyPI trailing month held at 97,003
- sol-pi: full rotation re-verification, citations intact; the 2026-11-05 push resolved non-default-branch with main quiet since 2026-10-18
- toon-format: pushed signal resolved non-material (CI-workflow commit unreleased, spec still v4.4), note untouched
- wrenai: pushed signal resolved non-material (unreleased 2026-11-02 SDK fixes, PyPI still 0.17.1), note untouched
- context-engines: matrix and index matched to the unchanged 21-member membership, front matter and internal links verified clean, no pivots

## (daily refresh, 2)

- control-planes: admitted Speakeasy Gram (speakeasy-api/gram, 277 stars, AGPL-3.0 stack behind the hosted AI control plane) on the coordinator handoff, superseding the 2026-10-10 closed-tier reading and the 2026-10-28 MCP-platform-kind deferral
- control-planes: admitted Agent Barn (aai-labs/agent-barn, 111 stars, Apache-2.0 self-hosted chat-coworker fleet plane with per-agent cost attribution and access roles) on the coordinator handoff, the borderline-traction call recorded with the zero-HN footprint stated
- control-planes: folded OktoLabsAI/okto-nexus into the okto-pulse note per the vendor-sibling ruling, recording single-winner handoff dispatch, eight harness runtime tools, and the R4 acceptance status with final G0 to G3 open
- speakeasy-gram: created from the coordinator handoff as the category's first usage-priced hosted gateway member, with druce.ai's governance-wiki entry as the critical source
- agent-barn: created from the coordinator handoff as the category's fifth live plane, with the OECD.AI catalogue entry as the independent coverage and the always-AUTO OpenClaw command approval as the stated caution
- okto-pulse: recorded the okto-nexus fold and refreshed counters (20 open issues, pushed 2026-11-02)
- agent-control: counters refreshed (51 open issues), the 2026-11-05 push identified as a side branch with the default branch quiet since v8.15.0
- agentfield: candidate line ran to v0.1.146-rc.17 (2026-11-04), PyPI holding at 0.1.145
- arifos: self-measurement stamp held at 2026-10-29 against a new repo_live_divergence drift line with deploy pending, PyPI still trailing at 1!2026.10.17
- cc-safety-net: npm train moved to 2.7.2 then 2.7.3 (env/sudo/command wrapper secret protection, remote-upload rule moved to Strict-only)
- cordum: counters refreshed (14 open issues), default branch quiet since 2026-10-28
- hol-guard: release train moved to v3.59.0 (2026-11-04), coverage contract still unrenewed 27 days past expiry
- ibm-contextforge: backlog refreshed (1,001 open issues, pushed 2026-11-04), v1.1.0 holding
- microsoft-agent-governance-toolkit: counters refreshed (134 open issues, pushed 2026-11-04), v4.1.0 holding
- mission-control: counters refreshed (59 open issues), v2.3.0 holding
- openappa: push date moved to 2026-11-04, v0.38.1 holding
- paperclip: backlog fell to 6,586 with the push date at 2026-11-05, release line holding at v2026.1023.0
- settlebridge: rotation re-verification found no drift, license metadata still unresolved
- sidjua: rotation re-verification found no drift, freeze banner and missed July reopen still up
- tinyagi: rotation re-verification found no drift, stall unchanged
- tracekit: release train moved to v1.5.3 (2026-11-03, PyPI in step)
- vestige: release train moved to v6.2.0 (2026-11-03), MCP surface fixed at six tools
- veto: counters refreshed (2 open issues), no new tag past veto-sdk@2.9.3
- control-planes: matrix extended to twenty-four columns with speakeasy-gram and agent-barn, columns re-sorted, and its Changes chronological order restored (the 2026-10-31 bullet sat after the 2026-11-01 one)
- agent-qa: recorded v0.2.6 (2026-11-03), npm tracks 0.2.6.
- crit: refreshed the activity line, default branch moved 2026-11-04, v0.23.2 still newest.
- deepeval: recorded Python 4.3.3 (2026-11-04), PyPI ahead of the GitHub release tags.
- hunk: refreshed the activity line, dependency security upgrades landed 2026-11-04 (PR 1256).
- jevals: refreshed the activity line, CI hardening push 2026-11-02, PyPI still 0.1.4.
- kiln: rotation re-verification, every cited URL live, activity line refreshed to the 2026-11-02 default-branch merge, v1.2.0 re-confirmed on GitHub and PyPI.
- kitaru: recorded the coordinated 2026-11-02 to 11-04 wave (PyPI 0.27.4 to 0.27.6, TypeScript 0.8.0, adapter-package releases).
- langfuse: recorded v4.73.0, v4.74.0, and v4.75.0 (2026-11-03/04) and refreshed the activity line.
- mlr-bench: rotation re-verification, every cited URL live, Hugging Face downloads refreshed with a fresh as-of date (82 to 118, rolling window, no release behind it).
- phoenix: refreshed the activity line, arize-phoenix-client 3.7.0 (2026-11-05), platform holds at 21.0.0.
- plannotator: recorded v0.28.43 through v0.28.49 (2026-11-01 to 11-04) and refreshed the activity line.
- pr-lens: refreshed the activity line, outside PR 89 merged 2026-11-04, npm CLI still 0.14.0.
- reviewbench: recorded the leaderboard-configuration disclosure merge (PR 109, 2026-11-04), no new onboardings past twenty.
- frontierharness-eval: rotation re-verification, every cited URL live, repository unchanged (pushed 2026-09-08, unlicensed, 82-point thread intact).
- evaluation-review: matrix maturity cells refreshed (agent-qa v0.2.6, Langfuse v4.75.0, Kitaru PyPI 0.27.6 and TypeScript 0.8.0, PR Lens age eleven weeks).
- evaluation-review: awesome-agent-observability scanned, two 2026-11-03 additions resolved out of category (Cognipeer/agent-sdk, recued/recued), routing flags filed.
- aeon: refreshed counters (297 forks, 1,542 commits, pushed 2026-11-04) and proof-of-work counts (151 findings, 113 repositories)
- airut: refreshed counters (pushed 2026-11-05); PyPI still 0.27.1, 287 trailing-month downloads
- claude-code-action: release train to v1.0.268 (2026-11-04), counters 2,246 forks and 856 open issues and pull requests
- claude-code-hooks: refreshed open issues to roughly 16.5k; 33-event table, five handler types, fail-open if filter, and the v2.1.267 and v2.1.295 gates re-confirmed on the live reference
- claude-code-routines: corrected three sibling claims against Copilot automations' shared type, re-verified out of rotation
- copilot-automations: full rotation re-verification; recorded personal versus shared automation types, the API trigger, AI-credit limits, and the billing split; creator-private caution now covers personal automations only
- executions: matrix trigger-types cell and intro, middle-ground, versioning, and choosing prose updated for Copilot automations' API trigger and shared types; no membership change
- github-agentic-workflows: recorded the two advisories published 2026-11-05 (high-severity compile-pipeline .tmp write, medium Copilot-token-to-api.openai.com, both patched in 1.2.0) and the v1.2.0 prerelease topping the train; forks 615
- n8n: full rotation re-verification, all claims held; Microsoft Teams joined the documented agent-channel list; counters 61.4k forks; pricing tiers unchanged
- rivet-agents: recorded the v0.5.5 release (2026-11-02); counters 138 forks, 89 open issues and pull requests
- trigger-dev: refreshed counters (1,525 forks, pushed 2026-11-03); v4.8.3 still the newest release
- executions: dedicated-source existence check found no credible executions list (sandbox lists already held by the sandboxing category)
- hybrid-execution: baml volatile re-verification: 0.23.1 nightly train through 20261103.b (11-04), no stable 0.23.1, forks 512, issues 370, npm month 1,162,193, pushed 11-05
- hybrid-execution: laya volatile re-verification: 0.4.21 through 0.4.24 released 11-01 to 11-05 (TS predictLong, vegaml core dependency, image/audio/video inputs), PyPI 0.4.24, pushed 11-05
- hybrid-execution: guardrails volatile re-verification: pushed 11-05 branch activity, default branch still the August hub-retirement merge, v0.11.0 latest
- hybrid-execution: instructor volatile re-verification: three 11-03 fixes joined the unreleased 1.18 line, pushed 11-03, downloads flat 8,516,567
- hybrid-execution: cua-s1 volatile re-verification: host pushed 11-05, component untouched since 10-20, spaces/driver/nightly releases not touching it
- hybrid-execution: kev volatile re-verification: pushed 11-03 branch activity, kev-1.0 latest, no new checkpoint
- hybrid-execution: marvin rotation re-verification: every cited URL holding, pushed 11-01 branch activity, train at v3.2.10
- hybrid-execution: ollaya rotation re-verification: decision2 reached the site, family count to thirty, 11-02 commits site polish, pushed 11-02, v0.18.0 latest, releases page 504 (API holding)
- hybrid-execution: strands-decider-2b volatile re-verification: four unreleased serving/deploy features (Lambda deploy, request warm-up, multi-GPU split, env benchmarks), pushed 11-04
- hybrid-execution: tool-eval-bench volatile re-verification: web UI run-type measurements feature #329 (11-03), v3.0.0 latest, pushed 11-03
- hybrid-execution: valen volatile re-verification: repository moved to Valen-Team/Valen, Valen-Decision-1.5 architecture merged 11-04, project renamed Valen-Decision in READMEs, chart and citations re-pointed, board ranks synced to the 10-29 regeneration
- hybrid-execution: jeeves rotation re-verification: every cited URL holding, weights 195 downloads (rolling swing, no trend read) and 8 likes
- hybrid-execution: outlines rotation re-verification: all seven references live, repository unchanged, vLLM docs redirect noted
- hybrid-execution: ollaya index one-liner moved to thirty families with the count change
- dex-horthy: out-of-rotation accuracy check confirmed the humanlayer handoff: repo README carried "pretty much all deprecated" from 2026-06-19 to the 2026-10-22 pivot, which open-sourced the rebuilt workspace monorepo (Apache-2.0, 12.8k stars as of 2026-11-05) and moved the HITL-era code to the legacy branch and legacy-final tag; What it is, Cautions, References updated and full re-verification run
- dex-horthy: episode #80 (Using AI for Writing, slated 2026-11-03) still flagged upcoming in the show repo as of 2026-11-05, aired count holds at 79 through #79 (2026-10-27)
- addy-osmani: rotation re-verification, all 15 cited URLs re-fetched and live, no claim moved (Anthropic MTS, Agentic Engineering book lead, Brownfield still the blog's latest, Substack archive ends at Agentic Verification 2026-10-14)
- boris-cherny: rotation re-verification, all 7 cited URLs re-fetched and live, GitHub followers moved to 11.8k as of 2026-11-05
- chip-huyen: rotation re-verification, 8 of 9 cited URLs re-fetched and live, no claim moved; Amazon listing 503 to fetches all run
- cole-medin: rotation re-verification, all 6 cited URLs re-fetched (YouTube surfaces JS shells only, as recorded 10-29), no claim moved
- dax-raad: rotation re-verification, all 12 cited URLs re-fetched and live, GitHub followers moved to 5.7k as of 2026-11-05, exoharness/exo authorship still excludes him
- people-and-publications: links, front matter, as-of dates, feature matrix (33 columns, sorted, cell-count parity), and category index (33 entries, alphabetical) verified clean; no membership change