← charleslikesdata.com afk cockpit
9 repos · 1,012 runs · 832 issues · 739 merged · data through Jun 7 – Aug 3, 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 = 751 gate-passed
98%
landed rate (executor repos)
n = 9 exec repos
$3.92
API-equiv $ / merged PR
n = 739 merged
fully loaded: includes failed runs, quarantines, and scout overhead
33%
quarantine (safety) rate
n = 832 issues
19%
escalation rate
n = 1,012 runs
0%
rejected rate
n = 1,581 proposals
Pipeline

Stage funnel

proposed
1,581
promoted
855
54% of proposed
attempted
832
97% of promoted
passed gate
751
90% of attempted
published
740
99% of passed gate
merged
739
100% of published
By repository

Outcomes per repo

landed quarantined other
Orbit-Wars
79 · 86%
afk-agent-system
439 · 90%
afk-cockpit
106 · 94%
breakroom
5 · 40%
graphmark
14 · 93%
household-bms
48 · 96%
household-command-center
1 · 100%
housing-commute-analysis
15 · 93%
the-vault
0 · 0%
the-workshop
125 · 81%
Producers

Proposal quality

dropped
296
harvest
5
mis-sized
88
pending
335
rejected
1
shipped
855
superseded
1
Safety axis

Quarantine by category

275 total
other
85
no-op
35
review
24
capability
21
integration-conflict
17
scope
17
build-crash
16
setup-crash
10
test-conflict
10
deny-surface
9
circuit-breaker
7
timeout
5
coupling
4
idle-timeout
4
environmental
3
unrecoverable-signature
3
flaky-test
2
security
2
rate-limit
1
Retries & recovery

Runs per issue

1.22 avg runs/issue
12345
Retries & recovery

Quarantine → retry rescue

retry rescues 60% of first-run quarantines · 106% of those actually retried
75%
first-try pass rate
16%
final-state quarantine (issues)
27%
quarantine rate (runs)
Cost · API-equivalent
fleet telemetry · per attempt · Jun 7 – Aug 3, 2026

Spend by model tier

$2,559.90 total across the window
sonnet$1666.58 · avg $1.69/attempt · n=984 attempts
unknown tier$670.16 · avg $1.69/attempt · n=397 attempts
opus$223.16 · avg $2.48/attempt · n=90 attempts
⚠ 26% of attempt spend ($670.16 · 397 attempts) has no recorded model_tier — telemetry gap, not a model
Overhead · orchestrator + scout ops
$338.79
301 cost events · additive to the tier spend above · not a model tier
Claude Code transcripts, afk worktrees only · per API call · Jul 4 – Aug 3, 2026
$0.03
median / call
$0.08
p90 / call
20,393
calls logged
whole machine, all projects: $8,945.16 across 59,426 calls
Cost dial

Per-API-call cost distribution

15124
4246
717
193
87
20
6
<$.05.05–.10.10–.15.15–.20.20–.30.30–.50$.50+
Model mix

Opus share by day

Jul 4 range: 0%–70% Aug 3
Throughput

Slices per day

1,012 runs
06-0706-1306-1906-2707-0307-1407-2307-30
Context cost

Tokens by type, per day

1.46B tokens · 95.6% cache reads — the context is overwhelmingly cached
cache read cache write input output