Skip to content

A local-first control layer for AI agents

active 2026-05-222026-07-15 (UTC)

Partial coverage24,437 / 26,316 hourly files (93%) · 2 absent upstream · 1,878 failed, retryable2023-08-152026-08-15 (UTC)— sampled evenly across the window, so rankings and trends hold; absolute counts scale up.
Events
681
Pushes
274
Pull requests
22
Issues
13
Stars
2
Forks
0

Activity over time

Daily event counts in the loaded window

Line chart, 55 days from 2026-05-22 to 2026-07-15. Pushes: 274 total, peak 21 in a day. Pull requests: 22 total, peak 2 in a day. Issues: 13 total, peak 4 in a day. Comments: 11 total, peak 2 in a day. Stars: 2 total, peak 1 in a day.

  • Pushes
  • Pull requests
  • Issues
  • Comments
  • Stars

Stars, PRs, issues and forks are under-captured in the later part of this window. GH Archive progressively stopped capturing non-push events during 2026 — −95% or worse by the end of the window. Every series here except Pushes fades for that reason, so a decline above reflects the archive, not this repository. Pushes stay reliable throughout, so read them, and the contributor counts derived from them, as the real signal. Data health has the measurements.

Top contributors

Pushes, PRs, issues, reviews and comments — stars and forks excluded, so this is contribution rather than popularity

ContributorContributionsPushesPRsComments
jerryfane3202742211

Recent activity

Latest issues, pull requests and releases

  • Issue comment#926jerryfane2026-07-14 15:44
    feat(dashboard-web): split workflow 'active' into active vs recent (#918)
  • Issue#738jerryfane2026-07-07 23:46
    skillopt train: review publish 422s on GitHub's 64KB issue-body limit + no local-review path in the state machine
  • Issue#721jerryfane2026-07-07 14:35
    claude Start() leaks the CLI JSON envelope as Raw — silently breaks skillopt synth/ab JSON consumers (bad_rubric, zero scores)
  • Issue comment#710jerryfane2026-07-07 13:28
    feat(workflow): preset prompt delivery modes full|referenced|auto
  • Issue#347jerryfane2026-07-07 13:27
    SkillOpt: offline rubric/metric induction from human feedback (AutoLibra-style)
  • Pull request#663jerryfane2026-07-05 04:06
  • Issue comment#658jerryfane2026-07-05 03:05
    Codex jobs never capture token usage — Deliver runs without --json, so budgets and dashboards undercount the largest runtime
  • Pull request#656jerryfane2026-07-05 01:49
  • Issue comment#652jerryfane2026-07-04 22:25
    spike(runtime): config-driven runtime metadata registry (NOT a plugin loader) — stop hardcoding model/effort/usage in the adapter switch
  • Pull request#653jerryfane2026-07-04 22:11
  • Releasejerryfane2026-07-02 19:20
    v0.8.3 — daemon CPU: pegged core → near-idle
  • Pull request#607jerryfane2026-07-02 10:56
  • Pull request#567jerryfane2026-06-30 23:29
  • Issue#550jerryfane2026-06-30 16:14
    daemon: singleton guard only covers 'daemon start', so 'daemon run' can silently run duplicate/orphaned daemons
  • Pull request#504jerryfane2026-06-27 10:24
  • Issue comment#502jerryfane2026-06-27 09:04
    watcher demo test (ignore, will close)
  • Issue comment#502jerryfane2026-06-27 08:53
    watcher demo test (ignore, will close)
  • Pull request#500jerryfane2026-06-27 07:33
  • Issue#487jerryfane2026-06-26 23:14
    Bug: --watch-issues Claude replies (@agent ask on issues) fail with 401 under contention, while direct agent ask --background succeeds
  • Issue#485jerryfane2026-06-26 22:51
    Review-agent follow-on: deterministic checkers (dup/lint/complexity/diff-size) as objective dimensions
  • Issue#483jerryfane2026-06-26 22:51
    Mode B follow-on: cross-family LLM-judge auto-pairwise (replace the human pick)
  • Issue#483jerryfane2026-06-26 22:51
    Mode B follow-on: cross-family LLM-judge auto-pairwise (replace the human pick)
  • Issue comment#476jerryfane2026-06-26 22:43
    Proposal (RFC): GitHub-backed agent templates — selective publish, backup & cross-machine restore
  • Releasejerryfane2026-06-25 17:40
    Gitmoot v0.6.0
  • Issue#427jerryfane2026-06-25 03:17
    Make Claude background-auth setup clearer (doctor/dashboard report the shell, not the daemon; token persistence undocumented)

Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 2 stars here means stars gained during the window, not the repo's star count.