Menu

Individual skill

tokens

0upvotes

Token-budget observability for godmode loops. Per-round cost tracking, token-log.tsv append, session-end summary, per-skill breakdown, delta vs prior sessions.

skills/tokens/SKILL.md

Read pinned source

Source commit 18bfc31d669804856ba232f04cdbd172afbdc379

What was checked

Matches the observed repository HEAD
Yes
Scanner rules are current
Yes
Repository scan complete
No
Skill scan complete
No
Observed behavior
Instructions only
Last checked
Nov 11, 2026

These are observations at the recorded commit, not a guarantee of safety. Review the source before running it.

Install this skill

Choose an agent in your Stack or copy a command below. Commands fetch the pinned commit and copy only this skill directory. Existing destinations are refused. Copying a command does not install anything.

An install command is unavailable: current rules, complete repository and skill scans, an eligible behavior level and a permitted license are required.

Repository context

Repository stars (not skill ratings)
28
Repository last updated
Aug 28, 2026
License
MIT

Observed agent integration: No specific integration evidence. Standard SKILL.md installation does not imply verified compatibility.

Files and excerpt

File metadata and permitted, redacted excerpts come from the recorded scan. Source links are pinned to that commit.

Browse files
Read excerpt

## Activate When
- `/godmode:tokens`, "token budget", "how expensive was this loop", "context cost"
- Any multi-round skill (optimize, refactor, debug, bench, terse) that runs a loop with DECIDE/LOG boundaries
- At session end, when the user asks "is my token usage trending up or down?"
- NOT for live streaming or real-time metering — this is append-per-round observability only. If you want a live meter, use a shell prompt, not this skill.
- Opt-out: if `GODMODE_TOKENS=0` is set in the environment, this skill is a no-op. Privacy-sensitive workflows (medical, secrets, PII sweeps) can set this to disable all round-size logging for the session. Opt-out is sticky per session; never overridden by auto-activation.

## Inputs
Ask once, cache for the session:
- `session_id` — stable id for the run. Default: `date +%s` at session start, reused across every round.
- `skill` — the skill driving the current round (read from session-state.json, not re-asked).
- `round` — monotonic counter, read from session-state.json.
- `terse_on` — boolean. `1` if `$GODMODE_TERSE` is non-empty, else `0`.
- `rtk_detected` — `1` if `command -v rtk` returns zero, else `0`. Logged as environment context, not used for computation.
Back to top