Skip to content

Mellea is a library for writing generative programs. From my work with the MIT-IBM Watson AI Lab.

active 2025-08-072026-08-07 (UTC)

Partial coverage11,656 / 12,831 hourly files (91%) · 2 absent upstream · 1,171 failed, retryable2025-02-222026-08-10 (UTC)— sampled evenly across the window, so rankings and trends hold; absolute counts scale up.
Events
2.9K
Pushes
609
Pull requests
253
Issues
247
Stars
171
Forks
34

Activity over time

Daily event counts in the loaded window

Line chart, 366 days from 2025-08-07 to 2026-08-07. Pushes: 609 total, peak 23 in a day. Pull requests: 253 total, peak 8 in a day. Issues: 247 total, peak 13 in a day. Comments: 715 total, peak 16 in a day. Stars: 171 total, peak 25 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
jakelorocco41415459109
planetf1221431675
nrfulton221291680
github-actions[bot]1681082238
avinash2692160842621
ajbozarth15022377
psschwei14732057
github-merge-queue[bot]13813800
mergify[bot]850085
HendrikStrobelt77241321
akihikokuroda450927
guicho271828405720
markstur320414
frreiss251611
AngeloDanducci22038
KenOcheltree16732
yelkurdi160110
araujof160210
mellea-auto-release[bot]111100
csbobby10036

Recent activity

Latest issues, pull requests and releases

  • Issue#1051ajbozarth2026-08-07 16:39
    feat: streaming span events
  • Pull request#1496github-actions[bot]2026-08-05 17:21
  • Issue comment#1417AngeloDanducci2026-07-20 19:17
    fix: force cpu accelerator for docling on apple silicon mps
  • Issue#1332AngeloDanducci2026-07-01 18:37
    refactor(intrinsics): guardian.py IOContract subclasses + Adapter constants (Epic #929 Phase 1 follow-up)
  • Issue comment#1353akihikokuroda2026-06-27 00:19
    fix: add Python code extraction from tool_calls in PythonCodeExtraction validator
  • Issue comment#996psschwei2026-06-23 21:47
    feat: context compaction strategies for the react loop
  • Issue#1273planetf12026-06-19 07:11
    bug: structured output via act() silently returns JSON string, not Pydantic instance
  • Issue#1217ajbozarth2026-06-18 17:47
    refactor(core): promote mot._thinking to public mot.thinking
  • Issue#1246psschwei2026-06-18 04:15
    Security hardening: webhook validation, adapter path traversal, API key hygiene, and input validation
  • Issue#641psschwei2026-06-18 04:12
    docs: Debug smolagents example incomplete response and serialization warning
  • Issue#454psschwei2026-06-18 04:12
    feat: self-repairing requirements
  • Issue#305psschwei2026-06-18 04:11
    Design: model routing during sampling
  • Issue#276psschwei2026-06-18 04:09
    Internationalization for `stdlib` components
  • Pull request#1280github-actions[bot]2026-06-17 08:33
  • Issue comment#1270AngeloDanducci2026-06-16 02:26
    feat: model aware sliding window context
  • Pull request#1263yehorcallmedai-maker2026-06-14 00:33
  • Pull request#1265github-actions[bot]2026-06-12 22:43
  • Pull request#1265akihikokuroda2026-06-12 22:43
  • Issue#1218ajbozarth2026-06-11 22:26
    fix(backends): unify raw-path token usage on mot.generation.usage; guard eval_count=None
  • Pull request#1254github-actions[bot]2026-06-10 22:38
  • Pull request#1243github-actions[bot]2026-06-09 23:28
  • Pull request#1243markstur2026-06-09 23:28
  • Issue comment#1239planetf12026-06-09 18:46
    ci: skip Python quality checks for docs-only PRs
  • Issue#1229ajbozarth2026-06-08 22:53
    bug(plugins): generation_post_call doesn't fire for pre-computed MOTs from _generate_from_context
  • Issue comment#1219psschwei2026-06-08 00:26
    fix(streaming): use public mot.cancelled property in streaming.py:437

Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 171 stars here means stars gained during the window, not the repo's star count.