This website requires JavaScript.
0adf7cdf2f
Merge pull request 'feat: add adaptive review effort budgets' (#19 ) from feat/review-budget-governor into main
masi
2026-09-01 12:15:16 +00:00
592747d98d
fix: resolve review budget feedback
2026-09-01 12:14:18 +00:00
4d51e14a1a
feat: adapt review budget to diff size
2026-09-01 11:50:46 +00:00
f9eef969e6
docs: document review budget defaults
2026-09-01 11:42:59 +00:00
3a10deef06
feat: add review effort accounting and budget governor
2026-09-01 11:41:08 +00:00
a202bd3598
Merge pull request 'refactor: split opencode runtime and tests' (#17 ) from refactor/split-opencode-modules into main
masi
2026-09-01 03:34:58 +00:00
e6cb7d01a6
fix: remove duplicate lens config helpers
2026-09-01 03:26:50 +00:00
f485d9faf9
refactor: split opencode runtime and tests
2026-09-01 02:54:23 +00:00
0c3997db35
Merge pull request 'refactor: organize pilot modules and tests' (#16 ) from refactor/split-pilot-modules into main
masi
2026-09-01 01:21:56 +00:00
8dfca67ad3
docs: describe review module seams
2026-09-01 01:17:00 +00:00
87dd695d97
refactor: split review pipeline responsibilities
2026-09-01 01:14:49 +00:00
c948f2818b
refactor: make review facade thin
2026-09-01 01:05:54 +00:00
7a510a926d
refactor: organize pilot packages
2026-09-01 00:59:51 +00:00
3a110ab52c
Merge pull request 'Refactor pilot architecture and remove dashboard' (#15 ) from refactor/pilot-architecture-langfuse into main
masi
2026-09-01 00:28:06 +00:00
9a9923d034
Merge pull request 'test: eval judges e2e' (#14 ) from test/eval-judges-trigger into main
masi
2026-09-01 00:26:59 +00:00
3f30bdd450
merge: resolve eval judge rule conflict
2026-09-01 00:26:25 +00:00
b4041f6892
refactor: split pilot architecture
2026-09-01 00:17:16 +00:00
5b8be61e2f
test: trigger review after DB rule insert
2026-08-31 19:19:59 +00:00
3109bd7c2d
pilot(eval): switch rule target from observation to trace
2026-08-31 19:18:59 +00:00
d746b1fdc2
test: trigger review for eval judges
2026-08-31 18:37:41 +00:00
370adcde6f
test: trigger after reopen
2026-08-31 18:36:07 +00:00
84325e61c1
test: trigger review after .pr-review.json merged to main
2026-08-31 18:34:18 +00:00
193a90e63e
Merge pull request 'feat(eval): LLM-as-judge evaluators, dataset item fixes, and Experiments runs' (#13 ) from fix/dataset-item-url-safe-ids into main
gitea_admin
2026-08-31 18:34:02 +00:00
1644c7f6b1
chore: enable pragent pilot on this repo (.pr-review.json on PR branch)
2026-08-31 18:33:06 +00:00
3543d15677
test: re-trigger after dedupe window
2026-08-31 18:31:32 +00:00
72ac0f76bc
test: retrigger review after eval rule wiring
2026-08-31 18:25:57 +00:00
5d44121b28
feat(eval): LLM-as-judge evaluators for finding actionability and review self-consistency
2026-08-31 17:17:16 +00:00
2e1ad817e7
feat(eval): filterable item metadata and dataset runs for the Experiments tab
2026-08-31 15:53:35 +00:00
1534a99630
fix(eval): dataset item ids that survive a URL path
2026-08-31 15:38:44 +00:00
d9eb4d9822
Merge PR #12 : langfuse evaluation layer
gitea_admin
2026-08-31 14:48:06 +00:00
2f96e66aab
feat(pilot): behavioural scorers, feedback ground truth, and an eval dataset
2026-08-31 14:22:55 +00:00
a4c35a4472
Merge pull request 'feat(pilot): emit per-review Langfuse traces' (#11 ) from feat/langfuse-tracing into main
gitea_admin
2026-08-31 13:03:14 +00:00
92283c44e8
feat(pilot): emit per-review Langfuse traces
2026-08-31 12:45:35 +00:00
51b81def98
feat(opencode): configure vllm-qwen38 provider for local Qwen3.8-27B
2026-08-24 22:49:50 +00:00
317942e15b
fix(pilot): strip opencode $schema + defensive harvest dict-strip
2026-08-24 19:38:48 +00:00
0d73a075cd
feat(opencode): bump qwen3.8-27b context limit to 57344
2026-08-22 19:21:53 +00:00
e5d4e252de
feat(opencode): multi-provider support with local provider for qwen3.8-27b
2026-08-22 19:13:06 +00:00
92020d4d46
feat(ai_review): route per-repo model override through PRICES provider field
2026-08-22 19:13:04 +00:00
e344831b05
feat(cost_model): add provider field on Price + register qwen3.8-27b (local)
2026-08-22 19:13:04 +00:00
2ba3ccbd95
docs(pilot): dashboard now behind Logto SSO
2026-08-22 18:05:02 +00:00
99255bb167
feat(dashboard): drop token auth, trust oauth2-proxy X-Forwarded-User
2026-08-22 18:03:33 +00:00
c9809bce9b
Merge deploy agent: k8s manifest + README
2026-08-22 17:01:34 +00:00
5fe83b0b2a
k8s(pilot): dashboard deploy manifest + README; pin NodePort 31540
2026-08-22 17:01:34 +00:00
218a8dc271
docs(pilot): dashboard README
2026-08-22 15:35:00 +00:00
e688ea61c5
Merge dashboard agent: data + HTTP server + mutation endpoints
2026-08-22 15:17:58 +00:00
dd49d83933
feat(dashboard): model select dropdown in edit form
2026-08-22 15:15:29 +00:00
6f7957010e
feat(dashboard): mutate .pr-review.json via Gitea contents API
2026-08-22 15:15:04 +00:00
be24ef245c
feat(dashboard): stdlib HTTP server with overview/repo/PR pages
2026-08-22 15:11:36 +00:00
bb9b6aa12d
feat(dashboard): read-only query module over feedback SQLite
2026-08-22 15:07:57 +00:00
fe5bebb4cf
Merge agent-B: cost-line fix + static_message + per-repo model override
2026-08-22 15:01:38 +00:00
d0f99b9763
feat(ai_review): parse .pr-review.json:model as per-repo override
2026-08-22 14:46:47 +00:00
f3a09b9397
Merge agent-A: move feedback*.py from WIP into tree
2026-08-22 14:46:46 +00:00
4e5f43ada7
feat(feedback): move feedback poster from WIP into pilot/
2026-08-22 14:46:07 +00:00
8472f35a58
feat(feedback): move feedback analyzer from WIP into pilot/
2026-08-22 14:46:07 +00:00
69e1fc06a2
feat(feedback): move feedback harvester from WIP into pilot/
2026-08-22 14:46:07 +00:00
6cdccb48ad
feat(feedback): move feedback storage layer from WIP into pilot/
2026-08-22 14:46:04 +00:00
7f37a36722
feat(ai_review): parse .pr-review.json:static_message + render as banner
2026-08-22 14:43:59 +00:00
b6b8173ccb
fix(ai_review): show actual model in cost line (was hardcoded glm-5.2:cloud)
2026-08-22 14:41:56 +00:00
a9e1b7ddfc
fix(opencode): patch apiKey from PRAGENT_MODEL_API_KEY at install_config
2026-08-22 13:12:46 +00:00
4c06a9ab3c
chore(opencode): swap glm-5.2:cloud → MiniMax-M2.7
2026-08-22 03:45:09 +00:00
d38e1c8693
test(opencode_review): skip posthash golden when feedback module absent
2026-08-22 01:43:05 +00:00
8e8ae54669
Merge feature/pragent-update: 16-task pragent update
2026-08-22 01:38:49 +00:00
661199dad2
chore(cost_model): drop stale AI-USAGE label references
2026-08-22 01:37:23 +00:00
3e03fb80a7
docs: onboarding uses .pr-review.json:enabled, not labels
2026-08-22 01:33:42 +00:00
2432228d68
feat(webhook): gate on .pr-review.json:enabled, drop labels
2026-08-22 01:30:07 +00:00
2c7d4803f1
feat(webhook): is_repo_enabled reads .pr-review.json:enabled
2026-08-22 01:24:10 +00:00
86352a3771
refactor(review): always render usage; drop report_usage flag
2026-08-22 01:18:34 +00:00
e1b74d982d
refactor(ai_review): drop AI_REVIEW/AI_USAGE label plumbing
2026-08-22 01:13:59 +00:00
33e4c16782
feat(usage): multi-provider equivalent cost table
2026-08-22 01:06:02 +00:00
979c93bdbb
feat(review): render walkthrough + risk_verdict + test_coverage
2026-08-22 00:58:51 +00:00
66628aae8d
fix(opencode_review): harden _synthesize_summary_fields (none-safe, dedupe emoji map)
2026-08-22 00:52:35 +00:00
e5a6e8923d
feat(opencode_review): python fallback for summary fields
2026-08-22 00:46:08 +00:00
2e982846d9
feat(review): walkthrough/risk_verdict/test_coverage schema
2026-08-22 00:36:50 +00:00
72b77a96e0
fix(confidence): pass multi_lens_observed kwarg to merge_confidence
2026-08-22 00:29:28 +00:00
f3a125666b
feat(ai_review): per-PR merge confidence 1-5 in header
2026-08-22 00:23:24 +00:00
9dae850887
test(config): fix cap-at-12 test to actually exercise the cap
2026-08-22 00:19:25 +00:00
0f7903377a
feat(config): parse enabled + compare_against
2026-08-22 00:12:06 +00:00
0b295e2443
fix(severity): extend badge label set + update SYSTEM_PROMPT
2026-08-22 00:06:22 +00:00
23ec1bf74a
feat(severity): add trivial + info levels
2026-08-21 23:58:57 +00:00
2cf4bdbfe8
feat(ai_review): render humanized token counts in usage + inline
2026-08-21 23:54:10 +00:00
3bcf825104
feat(ai_review): fmt_tokens() humanizes token counts
2026-08-21 23:46:24 +00:00
67339da8d0
feat(cost-model): add GPT, Gemini, Grok prices
2026-08-21 23:42:07 +00:00
99014a4cd3
chore(gitignore): exclude .worktrees/
2026-08-21 23:39:44 +00:00
f52b8d7803
docs(plans): 2026-08-21 pragent update implementation plan
2026-08-21 23:39:11 +00:00
f9f6ab4bf0
docs(plans): 2026-08-21 pragent update design
2026-08-21 23:32:49 +00:00
cdf116ece9
Merge PR #10 : read the AI-USAGE opt-in at render time
gitea_admin
2026-08-20 23:42:36 +00:00
c58c00181e
fix(review): read the AI-USAGE opt-in at render time, not at review start
2026-08-20 23:35:04 +00:00
aedea973ab
fix(webhook): label re-read must hit /api/v1, not the bare host
2026-08-20 23:29:03 +00:00
4ef62f28bb
fix(webhook): re-read PR labels at review start so a late AI-USAGE counts
2026-08-20 23:24:54 +00:00
7e4fd1975d
Merge PR #9 : diff compression + prior-review compaction + multi-lens orchestration
gitea_admin
2026-08-20 23:05:31 +00:00
2b1cf750b7
fix(review): correct diff-compression line numbers, prior-review dedupe, triage skip
2026-08-20 23:05:03 +00:00
78bcf6a9a0
feat(review): multi-lens orchestration — 5 parallel opencode subprocesses (security/docs/code-quality/tests/perf)
2026-08-20 22:16:21 +00:00
e8ebc54362
feat(review): ADDITIONAL_CONTEXT_URL — repo-provided static context cached per review
2026-08-20 17:45:51 +00:00
998f793ec2
feat(agent): tighten prompt to bound beyond-diff reads + de-generalize cost-model labels
2026-08-20 17:35:06 +00:00
f6be2b3c61
feat(review): PR-level collapsible metadata + emoji-tagged inline comments
2026-08-20 16:51:20 +00:00
5302e8dcd7
fix(review): salvage findings from nested-object fences + bare arrays + unfenced tail JSON
2026-08-20 16:29:44 +00:00
ec26ec000a
feat(review): equivalent provider price + enriched .pr-review.json schema
2026-08-20 16:12:46 +00:00
770581bf53
feat(input): add diff_compress module + prior-review compaction helpers
2026-08-20 16:12:33 +00:00
f59b906395
chore: scrub private infrastructure for a public repo, rewrite README
2026-08-18 12:59:52 +00:00
46513585ae
Merge PR #7 : harden the pilot against hostile PR content, add review skills + cost model
gitea_admin
2026-08-18 05:31:07 +00:00
masi
Claude
gitea_admin