← charleslikesdata.com afk cockpit
8 repos · 968 runs · 792 issues · 704 merged · data through Jun 7 – Aug 2, 2026

Autonomous-PR outcomes across the fleet. Cost is API-equivalent (notional on a Pro subscription). Quarantines are a safety signal — afk refused to ship — not failures.

Fleet state · not yet wired up
Live gate/promote/merge and dependency-frontier state isn't ingested yet — this panel is a placeholder for a future feature, not a fabricated number.
98%
landed rate
n = 716 gate-passed
98%
landed rate (executor repos)
n = 8 exec repos
$3.89
API-equiv $ / merged PR
n = 704 merged
fully loaded: includes failed runs, quarantines, and scout overhead
33%
quarantine (safety) rate
n = 792 issues
19%
escalation rate
n = 968 runs
0%
rejected rate
n = 1,487 proposals
Pipeline

Stage funnel

proposed
1,487
promoted
822
55% of proposed
attempted
792
96% of promoted
passed gate
716
90% of attempted
published
705
98% of passed gate
merged
704
100% of published
By repository

Outcomes per repo

landed quarantined other
Orbit-Wars
77 · 87%
afk-agent-system
412 · 89%
afk-cockpit
106 · 94%
graphmark
14 · 93%
household-bms
48 · 96%
household-command-center
1 · 100%
housing-commute-analysis
15 · 93%
the-vault
0 · 0%
the-workshop
119 · 80%
Producers

Proposal quality

dropped
293
harvest
5
mis-sized
80
pending
285
rejected
1
shipped
822
superseded
1
Safety axis

Quarantine by category

263 total
other
84
no-op
35
review
23
capability
21
integration-conflict
17
build-crash
16
scope
12
setup-crash
10
test-conflict
10
circuit-breaker
7
deny-surface
7
timeout
5
coupling
4
idle-timeout
4
unrecoverable-signature
3
flaky-test
2
security
2
environmental
1
Retries & recovery

Runs per issue

1.22 avg runs/issue
12345
Retries & recovery

Quarantine → retry rescue

retry rescues 61% of first-run quarantines · 104% of those actually retried
74%
first-try pass rate
16%
final-state quarantine (issues)
27%
quarantine rate (runs)
Cost · API-equivalent
fleet telemetry · per attempt · Jun 7 – Aug 2, 2026

Spend by model tier

$2,420.07 total across the window
sonnet$1544.28 · avg $1.64/attempt · n=940 attempts
unknown tier$670.16 · avg $1.69/attempt · n=397 attempts
opus$205.63 · avg $2.45/attempt · n=84 attempts
⚠ 28% of attempt spend ($670.16 · 397 attempts) has no recorded model_tier — telemetry gap, not a model
Overhead · orchestrator + scout ops
$317.55
285 cost events · additive to the tier spend above · not a model tier
Claude Code transcripts, afk worktrees only · per API call · Jul 3 – Aug 2, 2026
$0.03
median / call
$0.07
p90 / call
29,874
calls logged
whole machine, all projects: $9,056.04 across 68,075 calls
Cost dial

Per-API-call cost distribution

22816
5861
886
188
97
21
5
<$.05.05–.10.10–.15.15–.20.20–.30.30–.50$.50+
Model mix

Opus share by day

Jul 3 range: 0%–62% Aug 2
Throughput

Slices per day

968 runs
06-0706-1306-1906-2707-0307-1407-2307-30
Context cost

Tokens by type, per day

2.19B tokens · 95.9% cache reads — the context is overwhelmingly cached
cache read cache write input output