froot

durable maintenance loops · reputation read-model

watching mseeks/revisionist, mseeks/vibe-themer, mseeks/glassbox, mseeks/froot, mseeks/ynab-agent · generated 2026-06-04 23:00 UTC · derived live, stored nowhere

github 5 PRstemporal 15 workflowsclickhouse 142 spans / 3d

Dependency-patch

npm + uv · scan → bump → CI → merge5/5 live5 proposed67% merged2 awaiting you

Is it alive?

mseeks/revisionist · last 6h ago · next in 18h
mseeks/vibe-themer · last 6h ago · next in 18h
mseeks/glassbox · last 6h ago · next in 18h
mseeks/froot · last 6h ago · next in 18h
mseeks/ynab-agent · last 6h ago · next in 18h

Track record · the reputation

5
proposed
2
merged
2
awaiting
1
closed
67%
merge rate
90 min
median time-to-merge

Merge rate is the Stage-1 signal — narrow to npm patch bumps by construction. It counts a human merge, not a confirmed good outcome: revert tracking is a later loop, so merge is not yet proof of success.

Verification · CI is the oracle

1
CI passed
0
CI failed
1
no checks
0
timed out
3
unknown

A real oracle reported on 1 of 2 bumps with a reading. ‘no checks’ means CI was absent — not a pass; never conflated.

Model judgment · the one model call

3
clean
1
risky
1
unknown
0
no verdict

The model’s only job is the changelog verdict; the spine proposes the bump either way.

Approval gate · what a human owns

packagewaitingpr
vitest5h ago#22
agentmail3h ago#2

Bumps · the detail

packagebumpverdictcistateopenedpr
agentmail0.5.2 → 0.5.3unknownopen3h ago#2
vitest3.2.5 → 3.2.6riskyopen5h ago#22
nuxt3.21.6 → 3.21.7cleanpassedmerged22h ago#21
esbuild0.24.0 → 0.24.2cleanabsentmerged1d ago#8
zod3.23.7 → 3.23.8cleanclosed2d ago#3

Failures · where the loop did not close

bumpkindreasonwhen
mseeks-glassbox-left-pad-1.0.1terminatedERESOLVE: peer conflict2d ago

Determinism review

the transitive ring · advisory2 repos3 reviewed2 hazards

Is it alive?

mseeks/froot · last 6m ago · next due now
mseeks/ynab-agent · last 6m ago · next due now

Track record · the reputation

3
reviewed
1
flagged
2
clean
2
hazards
2
repos covered

The transitive ring: it chases first-party helper calls out of each workflow to catch a hazard the lexical CI kernel can’t see. Advisory — the blocking gate stays the kernel’s CI check. The hazard-resolved rate (was a flag gone on a later commit?) is a later loop; it needs accumulated history.

Reviews · the detail

repoprheadfindingsreviewed
mseeks/ynab-agent#1801ece6clean12m ago
mseeks/ynab-agent#26651455clean12m ago
mseeks/froot#12deadbee2 hazards datetime.datetime.now, random.random comment12m ago

Run telemetry

ClickHouse · trace-derived, best-effort142 spans / 3d

142 spans · 2 errored · last activity 12m ago · 3-day window.

activityrunsavgmax
open_pull_request1114162 ms55828 ms
analyze_pr92310 ms8140 ms
scan_candidates74100 ms9200 ms