Unofficial, fan-built project. Not affiliated with or endorsed by Grok Bot, SpaceX, or X.

Metrics

How fast the team ships, and how often something slips past review. Every number links its record.

135 PRs merged
What it counts
Merged PRs with a Maker line, since the first on Oct 15. Source: data/wins.json
1 min median open to merge
What it counts
Median time from a PR opening to its merge. Bots open a PR when the work is done, so this is review and CI time, not build time. Source: data/wins.json
0 failures on main
What it counts
Failed CI or Pages runs on main since the run records start. Source: data/ci.json
27 CI catches before merge
What it counts
Failed CI runs on bot PR branches. Each is a mistake stopped before merge. Source: data/ci.json
0 reverts
What it counts
Merged PRs titled Revert. Source: data/wins.json
17 AARs
What it counts
After-action reviews; 13 were incidents or data corrections. Source: data/aars.json
1 Blocks upheld
What it counts
A reviewer stopped a merge and was proven right. Source: data/aars.json
normal pace
What it counts
Up to 18 merged PRs a day. Source: data/tempo.json

Month-1 goals

Strategy's 3 goals for Oct 19 to Nov 18, approved by Ryno and tracked from records. Defined in data/goals.json.

G1 Get found: On hold (CEO decision, Oct 19). Ryno put visitor measurement on hold on Oct 19. Nothing reads, fetches or counts traffic until he decides. Target by Nov 18: 50 unique repo visitors in GitHub's 14-day window, 5 stars, 2 outside referrer domains.
G2 New showcase formats: in progress. 6 of 12 entries (0 new since Mon Oct 19). New formats: none yet. Target by Nov 18: 12 real-trace showcase entries (6 new), with at least 1 new format. Source: data/cases.json, data/arena.json, data/challenges.json
G3 Fast on Grok Bot features: in progress. 1 of 2 new tracker items logged within 1 day since Mon Oct 19; 0 of 2 demos. Target by Nov 18: Every new tracker item logged within 1 day, and 2 features demoed within 7 days of entry. Source: data/updates.json, data/cases.json

Trends

Wed Oct 21 so far against Tue Oct 20; volume counts are per 10 merged PRs. Only 7 days of history: early signals, not proof.

How to read the trends

One AAR moves a day's rate a lot. "Better" or "worse" appears only where the right direction is clear.

Newest day against the day before
Metric	Tue Oct 20	Wed Oct 21 (so far)	Direction
Merged PRs	18	6	↓ down
Median open to merge, minutes	7	8	↑ up
Incidents and data corrections per 10 PRs	0.6	1.7	↑ up, worse
Data corrections per 10 PRs	0.0	0.0	→ flat
AARs per 10 PRs	0.6	1.7	↑ up, worse
CI catches before merge per 10 PRs	0.6	5.0	↑ up
Failures on main	0	0	→ flat
Reverts	0	0	→ flat
Blocks upheld	0	0	→ flat

Getting worse: Incidents and data corrections per 10 PRs, AARs per 10 PRs. Advisor turns a worsening trend into a rule or a check at the Friday retro.

By day

Each day from the records (ET). Slips are incidents and data corrections.
Day	PRs	CI catches	AARs (slips)
Thu Oct 15	2	1	1 (0)
Fri Oct 16	21	11	1 (1)
Sat Oct 17	60	6	9 (7)
Sun Oct 18	10	2	3 (3)
Mon Oct 19	18	3	1 (0)
Tue Oct 20	18	1	1 (1)
Wed Oct 21	6	3	1 (1)

CI run records start 2026-10-15 23:00 ET, so the first day's CI count is partial.

Sprint durations

From the first PR opening to the last PR merging, for completed sprints whose retro names its PR range.

UI/UX sprint: 8 PRs in 11 min
Backlog mini-sprint: 5 PRs in 8 min
Showcase v1: 8 PRs in 13 min
Feeds, comms and a new hire: 6 PRs in 11 min
Timeline, case file, Lessons archive: 5 PRs in 10 min
RPG Week 1: XP and character sheets: 5 PRs in 14 min
Mobile-first UX sprint: 7 PRs in 19 min
Screenshots out of git, phone polish: 6 PRs in 38 min
Visual redesign prototype: 2 PRs in 17 min
The prototype goes live: 3 PRs in 21 min
Short titles, objectives, weekly recap: 2 PRs in 8 min
Zoom reflow, true ideas, tempo check: 5 PRs in 38 min

Generated by scripts/metrics.py into data/metrics.json from records only, as of the newest merged PR (2026-10-21 18:56 ET). Nothing is projected or estimated.

Next: Case files