← charleslikesdata.com afk cockpit
6 repos · 745 runs · 602 issues · 526 merged · data through Jun 7 – Jul 10, 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.
99%
landed rate
n = 533 gate-passed
99%
landed rate (executor repos)
n = 5 exec repos
excludes my-brain (10 quar)
$3.62
API-equiv $ / merged PR
n = 526 merged
fully loaded: includes failed runs, quarantines, and scout overhead
37%
quarantine (safety) rate
n = 602 issues
21%
escalation rate
n = 745 runs
0%
rejected rate
n = 900 proposals
Pipeline

Stage funnel

proposed
900
promoted
614
68% of proposed
attempted
602
98% of promoted
passed gate
533
89% of attempted
published
527
99% of passed gate
merged
526
100% of published
By repository

Outcomes per repo

landed quarantined other
Orbit-Wars
51 · 98%
afk-agent-system
312 · 99%
afk-cockpit
106 · 99%
claude-workflow
75 · 98%
household-bms
48 · 100%
my-brain
10 · 0%
Producers

Proposal quality

dropped
151
harvest
5
mis-sized
32
pending
96
rejected
1
shipped
614
superseded
1
Safety axis

Quarantine by category

221 total
other
80
no-op
38
capability
17
review
17
build-crash
16
integration-conflict
14
setup-crash
10
test-conflict
6
deny-surface
5
idle-timeout
4
timeout
4
scope
3
coupling
2
flaky-test
2
unrecoverable-signature
2
security
1
Retries & recovery

Runs per issue

1.24 avg runs/issue
12345
Retries & recovery

Quarantine → retry rescue

retry rescues 60% of first-run quarantines · 103% of those actually retried
71%
first-try pass rate
17%
final-state quarantine (issues)
30%
quarantine rate (runs)
Cost · API-equivalent
fleet telemetry · per attempt · Jun 7 – Jul 10, 2026

Spend by model tier

$1,748.71 total across the window
sonnet$938.14 · avg $1.36/attempt · n=692 attempts
unknown tier$674.64 · avg $1.69/attempt · n=399 attempts
opus$135.93 · avg $2.34/attempt · n=58 attempts
⚠ 39% of attempt spend ($674.64 · 399 attempts) has no recorded model_tier — telemetry gap, not a model
Overhead · orchestrator + scout ops
$155.00
130 cost events · additive to the tier spend above · not a model tier
Claude Code transcripts, afk worktrees only · per API call · Jun 10 – Jul 10, 2026
$0.03
median / call
$0.08
p90 / call
40,491
calls logged
whole machine, all projects: $6,267.97 across 64,483 calls
Cost dial

Per-API-call cost distribution

32197
6268
1579
311
106
27
3
<$.05.05–.10.10–.15.15–.20.20–.30.30–.50$.50+
Model mix

Opus share by day

Jun 10 range: 0%–72% Jul 10
Throughput

Slices per day

745 runs
06-0706-1106-1506-1906-2506-2907-0307-10
Context cost

Tokens by type, per day

2.61B tokens · 95.3% cache reads — the context is overwhelmingly cached
cache read cache write input output