AI pilot licence

Every flight, on the record: the work I ship with Claude as co-pilot, counted from the session logs. None of it is self-assessed.

Totals cover the period since 8 March 2026, through 30 October 2026, read from the durable activity record. That record has been kept since 7 July 2026. From before then it holds only the sessions whose logs were still on disk, so the earlier months are incomplete. A message is one transcript record: a prompt, a tool result or one block of a model reply.

AI pilot licencechecking

Federation of AI Aviation

AIP-2026-5152

Bradley Isenbek

Airline transport · Multi-engine (10 models) · Instrument rated (8 ratings)

ClassATP

Issued: 8 March 2026
Expires: NEVER
Last entry: 30 October 2026

Sessions: 5,152; flights logged
Messages: 1,451,267; transcript records
Tool calls: 456,035; made by the model
Projects: 132; directories flown from
Models flown: 10; type ratings below
Active days: 215; of 237 in the period

ai-pilot-data.json license and streaks, 8 Mar to 30 Oct 2026, generated 30 October 2026

Class ATP is a volume band: more than 300,000 transcript records. The serial is the session count, the issue date is the first day in the record, and the two lines at the foot repeat the card in machine-readable form. In flight means a Claude Code session on this host wrote to its log in the last five minutes; sessions on my second machine are not seen from here.

Logbook

The last 24 hours first, counted a minute at a time. Then the whole record: every day since 8 March 2026, with each week's total over its column, and the hours of the day the work lands in.

Claude Code, last 24 hourslast heard 30 Oct 2026 14:29 UTC

778 min active in the last 24 h

15 so far this hour, 42 in the hour before. Claude Code sessions on this host only.

Active minutes per clock hour (UTC)

tallest 56 min

Active minutes per clock hour (UTC)
Bucket	min
29 Oct 14:00 UTC, part hour	19
29 Oct 15:00 UTC	27
29 Oct 16:00 UTC	22
29 Oct 17:00 UTC	27
29 Oct 18:00 UTC	52
29 Oct 19:00 UTC	56
29 Oct 20:00 UTC	45
29 Oct 21:00 UTC	25
29 Oct 22:00 UTC	36
29 Oct 23:00 UTC	27
30 Oct 00:00 UTC	37
30 Oct 01:00 UTC	33
30 Oct 02:00 UTC	55
30 Oct 03:00 UTC	35
30 Oct 04:00 UTC	22
30 Oct 05:00 UTC	27
30 Oct 06:00 UTC	22
30 Oct 07:00 UTC	21
30 Oct 08:00 UTC	22
30 Oct 09:00 UTC	20
30 Oct 10:00 UTC	22
30 Oct 11:00 UTC	33
30 Oct 12:00 UTC	36
30 Oct 13:00 UTC	42
30 Oct 14:00 UTC, hour in progress	15

Counted once a minute by a cron on this host. Last count 30 Oct 2026 14:29 UTC

public/data/activity-pulse.json: rewritten every minute, this host only

Every day on the record215 active days of 237

Current streak, the longest: 156 days; 28 May to 30 Oct, every day
Busiest day: 24,524; records on 12 October 2026
Busiest week: 105,773; records, week of 12 October 2026

Transcript records per day and per calendar week, 8 March 2026 to 30 October 2026

tallest week 105,773 records

One cell is one UTC day, 2026-03-08 to 2026-10-3001 to 650651 to 4,2714,272 to 8,1518,152 to 11,33711,338 to 24,524recordsbusiest day

Transcript records per day and per calendar week, 8 March 2026 to 30 October 2026, by month
Month	records	Active days	Days in window
2026-03	1,846	15	24
2026-04	14,057	24	30
2026-05	25,226	24	31
2026-06	161,610	30	30
2026-07	245,101	31	31
2026-08	350,793	31	31
2026-09	288,261	30	30
2026-10	364,373	30	30

A streak is consecutive UTC days with at least one record. The bar over each column is that calendar week, Monday to Sunday. The first and last are part weeks. The pale months at the start are not a slow start: the record has been kept since 7 July 2026, and before that it holds only the sessions whose logs were still on disk.

Hours of the dayAmerica/Detroit

Transcript records by hour of day (America/Detroit)

tallest 117,776 records

Transcript records by hour of day (America/Detroit)
Bucket	records
00:00 America/Detroit	43,778
01:00 America/Detroit	32,348
02:00 America/Detroit	15,540
03:00 America/Detroit	8,071
04:00 America/Detroit	4,635
05:00 America/Detroit	4,318
06:00 America/Detroit	6,169
07:00 America/Detroit	9,856
08:00 America/Detroit	39,931
09:00 America/Detroit	57,381
10:00 America/Detroit	78,488
11:00 America/Detroit	87,349
12:00 America/Detroit	91,539
13:00 America/Detroit	102,694
14:00 America/Detroit	113,236
15:00 America/Detroit	117,776
16:00 America/Detroit	103,652
17:00 America/Detroit	98,558
18:00 America/Detroit	77,021
19:00 America/Detroit	62,741
20:00 America/Detroit	72,769
21:00 America/Detroit	75,893
22:00 America/Detroit	81,170
23:00 America/Detroit	66,354

Sessions started by hour of day (America/Detroit)

tallest 241 sessions

Sessions started by hour of day (America/Detroit)
Bucket	sessions
00:00 America/Detroit	201
01:00 America/Detroit	200
02:00 America/Detroit	205
03:00 America/Detroit	196
04:00 America/Detroit	197
05:00 America/Detroit	195
06:00 America/Detroit	196
07:00 America/Detroit	195
08:00 America/Detroit	197
09:00 America/Detroit	219
10:00 America/Detroit	217
11:00 America/Detroit	227
12:00 America/Detroit	236
13:00 America/Detroit	224
14:00 America/Detroit	233
15:00 America/Detroit	238
16:00 America/Detroit	241
17:00 America/Detroit	221
18:00 America/Detroit	228
19:00 America/Detroit	214
20:00 America/Detroit	222
21:00 America/Detroit	225
22:00 America/Detroit	207
23:00 America/Detroit	218

Hours of the day in America/Detroit, summed over the whole period. The bar marked now is the hour it is there as you read this.

ai-pilot-data.json activityHeatmap, hourlyDistribution and streaks, 8 Mar to 30 Oct 2026, generated 30 October 2026

Models flown

10 models, each drawn from the first day to the last day it appears in the record. The figure on the right is its share of all output tokens, and the type rating is a band of that share and nothing more: Expert above 50%, Proficient above 20%, Familiar above 5%, Exposure below that.

Type ratingsshare of output tokens

When each model was flown, 8 March 2026 to 30 October 2026

Opus 5Proficient · 26 Jul to 29 Oct · 142,152 responses29.5%

Opus 4.8Proficient · 28 May to 16 Oct · 95,764 responses26.9%

Fable 5.1Familiar · 1 Sep to 30 Oct · 59,083 responses15.5%

Opus 5.5Familiar · 23 Sep to 30 Oct · 78,076 responses13.0%

Fable 5Familiar · 10 Jun to 23 Sep · 58,885 responses12.8%

Opus 4.7Exposure · 17 Apr to 22 Jul · 10,388 responses1.9%

Opus 4.6Exposure · 8 Mar to 27 Apr · 2,239 responses0.3%

Sonnet 5Exposure · 1 Jul to 26 Aug · 172 responses<0.1%

Sonnet 4.6Exposure · 2 Apr · 215 responses<0.1%

Haiku 4.5Exposure · 14 Jul · 1 response<0.1%

Mission log

A mission is a project, and a project is the directory a session started in, so these are directory names as they sit on disk. Each bar runs from the first active day to the last. The figure on the right is messages. Subagent logs are not counted.

Missionsbusiest 20 of 132 projects

First to last active day, and messages, 8 March 2026 to 30 October 2026

cb*, 13 directories116 sessions · 13 Apr to 30 Oct533,902

infra12 sessions · 6 Jun to 30 Oct121,563

terrapulse23 sessions · 29 May to 26 Oct97,223

65025 sessions · 12 Aug to 30 Oct69,580

context-agent4,719 sessions · 11 Oct to 30 Oct63,942

public1 session · 22 Aug to 29 Oct48,075

bradleyio7 sessions · 8 Mar to 30 Oct46,709

The busiest 20 of 132 projects by messages. The row marked cb* is 13 directories whose names begin with cb, summed. Most are repositories in the Nominate-AI organisation, which are private. /work gives that organisation's totals without naming its repositories, and so does this log. One more is not listed at all; its sessions and messages are still in every total on this page.

ai-pilot-data.json typeRatings and missionLog, 8 Mar to 30 Oct 2026, generated 30 October 2026

Top tools

What the model reached for. 456,035 tool calls in the period, across 144 distinct tools. Most of flying is the shell.

Toolsshare of all calls

The next 8, share of all tool calls

ToolSearch1.5%

mcp__tinyagent__report1.4%

mcp__tinyagent__stamp1.3%

SendMessage1.1%

mcp__tinyagent__due1.0%

TaskUpdate0.7%

Agent0.4%

Monitor0.4%

Shares, not counts, on purpose. This list is read from the raw tool-call table, which holds some rows twice (it runs about 1.4% over the corrected total above). The duplicates are spread evenly enough that the shares hold, but the counts would not.

pilot-analytics.json by_tool, 8 Mar to 30 Oct 2026, generated 30 October 2026, the raw tool-call table of the same activity record

Ratings

Read these as rough. They measure what the work touched, not how well it went. The instrument ratings describe about the last month, not the whole record.

How these are scored

Instrument ratings come from the words in each project's CLAUDE.md, the plan files and the technologies named in the session logs still on disk, which reach back about 30 days. Each domain's score is half keyword coverage and half its mentions against the busiest domain's, and keywords are matched as plain substrings.

Competency scores are fixed formulas over counts from the record, the plan files on disk and the prompt history, and the count each one was computed from is printed under its name. Debugging assumes every shell, search and read call is debugging, which overstates it. The formulas stop at 95, and 3 scores sit at that cap. Piloting style is two ratios: directive is twice the share of tool calls that were Edit or Write, and plan first rises with plan files per session.

Ratings0 to 100

Instrument ratings, by domain

AI / ML3,071 mentions · 63% of its keywords81

Backend1,839 mentions · 83% of its keywords71

Security760 mentions · 100% of its keywords62

Systems424 mentions · 100% of its keywords56

IoT / Edge754 mentions · 74% of its keywords49

Frontend568 mentions · 77% of its keywords47

DevOps513 mentions · 78% of its keywords47

Data Engineering641 mentions · 68% of its keywords44

Competency profile

Tool Mastery144 tools, 456,035 calls95

Debugging357,610 Bash, Grep and Read calls95

Multi-Project38,670 project switches in the prompt history95

Architecture132 projects70

Iteration281 records/session avg62

Planning1 plan files on disk44

Piloting style: Balanced

Collaborativeagainst 23% directive77%

Plan firstagainst 50% iterate50%

What came up

Technologies by how often they appear in a sample of the session logs still on disk (100 files), which reach back about 30 days, and in each project's CLAUDE.md. This counts mentions, not proficiency, and it describes recent work, not the whole record.

Skills cloudtop 12 of 25

Mentions in session transcripts

JSON2,292

Nginx1,577

Python1,220

Bash1,134

SSH899

HTML874

CSS722

SQLite601

systemd526

DuckDB451

Anthropic441

YAML341

Go, Git, Rust, SPI, AWS are left out. The count matches a name anywhere inside a word, and those names turn up inside ordinary English (going, digit, trust, despite, draws), so their totals are not mentions of the technology.

ai-pilot-data.json competencyRadar and pilotingStyle, 8 Mar to 30 Oct 2026; instrumentRatings and skillsCloud, a sample of the logs still on disk (about the last 30 days), generated 30 October 2026

Token economy

What the flying burned. Nearly all of it is cache reads: the same context, read again on every turn.

Tokenssince 8 March 2026

Token totals, 8 March 2026 to 30 October 2026
Input	14.2M14,208,799
Output	399.9M399,888,252
Cache read	146.4B146,355,115,152
Cache created	2.4B2,358,758,539
Cache efficiency	98.4%of the input side
Model responses	446,975
Cost at API list price	not computed

Input plus output tokens per day, last 30 active days

tallest 5,600,568 tokens

Input plus output tokens per day, last 30 active days
Bucket	tokens
1 October 2026	2,156,095
2 October 2026	2,192,436
3 October 2026	814,767
4 October 2026	1,465,813
5 October 2026	2,530,422
6 October 2026	3,940,592
7 October 2026	4,763,530
8 October 2026	2,374,125
9 October 2026	1,943,564
10 October 2026	1,135,469
11 October 2026	3,063,082
12 October 2026	5,600,568
13 October 2026	2,037,163
14 October 2026	2,335,824
15 October 2026	4,335,325
16 October 2026	4,181,318
17 October 2026	1,491,845
18 October 2026	2,338,320
19 October 2026	2,400,408
20 October 2026	3,252,732
21 October 2026	2,816,251
22 October 2026	3,228,140
23 October 2026	3,916,388
24 October 2026	2,181,558
25 October 2026	1,270,161
26 October 2026	3,203,857
27 October 2026	3,357,286
28 October 2026	4,002,538
29 October 2026	4,031,446
30 October 2026	614,604

Cache tokens are not in these bars. The last bar is the day the file was generated, so it is a part day.

ai-pilot-data.json tokenEconomy, 8 Mar to 30 Oct 2026, generated 30 October 2026

Each total is shown abbreviated and, under it, exactly. Cache efficiency is cache read as a share of everything on the input side: input, cache read and cache created together. Cost is not computed: the durable record does not say whether cache writes used the 5 minute or the 1 hour TTL, which are priced differently.

How tokens are counted. Counted once per model response, not once per log line. The record does not store response ids, so a response is reconstructed from the usage counts its log lines repeat. On 2026-10-02 that rule was compared with response ids in the raw logs still on disk, about 40 percent of the record, and gave identical counts and token sums. The other 60 percent, whose logs were already pruned, is counted by the same rule and was not verified. Tokens used by subagents are not in the record.

Keep reading

This page is the receipts. Two longer reads sit beside it: The shift, the argument, and Cost analysis, the bottom-up cost model.