Repository navigation
Releases
v0.1.1 - one answer for the same window
The prose review found a real defect: two tools were reporting different totals for the
same window.
Fixed
Three tools, one answer. cf-cost summed per file, so a resumed session was charged
once per replay — $3,566 / 37 sessions against cf-analyze's $3,562 / 34. The page counted a
fork's replayed boundary records again, so it said 120 boundaries / 19.4M dropped where the CLI
said 119 / 19.2M; its error table was inflated the same way. One question — what counts as one
session? — was answered in three places. dedupeResumed now lives in lib/transcript.mjs and
all three import it; the atlas also dedupes at the record level, because a fork carries a new
session id and the id-level rule cannot see the replay.
Verified on one machine's logs, 30-day window:
| tool | total | boundaries | dropped | errors |
|---|---|---|---|---|
cf-analyze |
$3,562.19 | 119 | 19,247,822 | 106 |
cf-cost |
$3,562.19 | — | — | — |
cf-atlas |
$3,562 | 119 | 19.2M | 106 |
The README quoted the wrong numbers — including one this changelog calls a bug. It said
399.6M dropped tokens, which is the figure the cumulative-counter defect produced. The same
wrong figure shipped inside read-sessions.zip. Restated with a measurement window.
Both READMEs said "four skills" after build-atlas shipped in v0.1.0; skills/README.md
omitted it from the table entirely. A test now fails if a skills/ directory is missing from
either README.
The test badge claimed a count nobody maintained — 149 when the suite was 169, then 170,
then 177. Replaced with the live CI badge.
Added
docs/release.md — how a zip is built, why CreateFromDirectory needs absolute paths (a
relative call silently produced five identical 22-byte empty zips), and the check that catches
a fix landing in only one tool: compare the packed copy's numbers, not just that it starts.
Assets
Five skill zips, SKILL.md at each archive root. Every entry was diffed byte-for-byte against
this tag before upload. The zips embed file mtimes, so they are not reproducible byte-for-byte —
what is reproducible is the content.
177 unit tests, green on 3 OSes × Node 20/22/24.
MIT · acrot0
v0.1.0 — Claude Code session forensics
Five skills for reading what Claude Code actually did.
Skills
| Skill | What it answers |
|---|---|
read-sessions |
Which model actually answered when you asked for another; where the tokens went; what compaction threw away |
audit-install |
Why a hook, skill, plugin or CLAUDE.md silently stopped working |
triage-failure |
What actually broke when a request failed with an empty error box |
audit-spend |
Where the spend went, per model and per prompt |
build-atlas |
One offline HTML page of the whole history |
Each ships as a zip with SKILL.md at the archive root. Install by unzipping into your skills directory, or copy the folder from skills/ in the repo.
Why the numbers can be trusted
Three defects were found by generating a real report and checking it against the raw logs, not by reading the code:
- Compaction loss was multiplied.
cumulativeDroppedTokensis cumulative per session; summing it over boundaries reported 437M dropped where the per-event deltas total 19.4M. - Resumed sessions were counted twice. A resumed session replays its history into a new file with the same cumulative cost. 398 files held 330 session ids; summing files turned ~$397 into $3,562.
- Spend was called an estimate when it was a placeholder. The client flags unknown-priced models itself, and 24 of 34 records carried that flag.
What is tested
- 169 unit tests (
npx vitest run --pool=forks) - 7 trigger evals (
claude plugin eval .) — recorded run 7/7 at--runs 1; the honest default is--runs 3and costs about three times as much
The limit worth knowing
Every figure rests on an undocumented file format. If Claude Code changes its shape, these tools report unmeasured and print no totals rather than printing zeroes. The report footer states the date the format was last verified, and the command to re-verify it.
Read-only. Node 20+. No runtime dependencies, no network, no telemetry.
MIT · acrot0