Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add Kolikkii/token-razor --skill minimize-token-usagegit clone --depth 1 https://github.com/Kolikkii/token-razorWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/kolikkii/token-razor/minimize-token-usage)<a href="https://agentmods.dev/skills/kolikkii/token-razor/minimize-token-usage"><img src="https://agentmods.dev/badge/skills/kolikkii/token-razor/minimize-token-usage/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/kolikkii/token-razor/minimize-token-usage"><img src="https://agentmods.dev/badge/skills/kolikkii/token-razor/minimize-token-usage.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00082 | $0.00565 |
| Opus 5 | $0.00041 | $0.00282 |
| Sonnet 5 | $0.00016 | $0.00113 |
| Haiku 4.5 | $0.00008 | $0.00056 |
Grade A, and why
minimize-token-usage scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 49 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Minimize Token Usage
Treat context as a budget. Preserve decisions, constraints, failures, and verification evidence; discard repetition and low-signal bulk.
Work economically
- Inspect narrowly with
rg, targeted line ranges, and exact files. - Reuse evidence already in the thread. Do not reread unchanged content.
- Batch independent reads only when every result is likely to matter.
- Prefer summaries, counts, and diffs over full listings.
- Keep progress and final prose proportional to the request.
- Verify risky changes; token savings never justify skipping relevant tests.
Handle compressed results
Token Razor labels a compacted tool result with a TR-XXXXXXXXXXXX identifier. Use the visible summary first. Restore the full result only when omitted evidence blocks the next decision:
node "/installed/token-razor/scripts/cli.mjs" restore TR-XXXXXXXXXXXX --data-dir "/plugin/data/path"
Copy the exact command from the compacted result; it contains the current installation and data paths. Prefer a fresh narrow query over restoring a huge archive when the missing fact is known. Never invent omitted content.
When the result advertises a search command, use it before printing the full archive:
node "/installed/token-razor/scripts/cli.mjs" search TR-XXXXXXXXXXXX "decisive literal" --context 3 --data-dir "/plugin/data/path"
Select a mode
balanced: default; compress large tool output aggressively while retaining signals.safe: larger output budget for unfamiliar or high-risk debugging.extreme: smallest budget for repetitive logs, broad searches, and explicit token minimization.passthrough: use only when the user asks for verbatim or unabridged output.
The prompt hook recognizes these intents per turn. Persistent project overrides belong in .token-razor.json; see algorithms-and-configuration.md only when tuning or auditing behavior.
Guard correctness
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 8d ago First seen · 49 lines · 82 tokens per session scan A 2e087bf0d56c
minimize-token-usage is a skill published in the GitHub repository Kolikkii/token-razor (1 stars, last pushed 1mo ago), licensed MIT. It adds 82 tokens to every session and 565 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
opensrc
Fetch dependency source code to give AI agents deeper implementation context. Use when the agent needs to understand how a library works internally, read source code for a package, fetch implementation details for a dependency, or explore how an npm/PyPI/crates.io package is built. Triggers include "fetch source for"…
nopus-configure
Configure nopus complexity sensitivity, extra-simple rewrites, rewrite evidence, and Pi response hiding for this user.
nopus-simplify
Rewrite the immediately preceding assistant response with clearer and more direct prose when the user invokes this skill.
pace-workflow
Use for PACEflow-managed coding work: create, approve, resume, verify, close, or archive CHG/HOTFIX before project file edits.
pace-bridge
Use to bridge a confirmed Superpowers/native plan into PACEflow CHG/HOTFIX artifacts, create artifact-writer prompts, and mark the specific plan as synced.
audit
Use for PACEflow internal full audit: run five independent review agents, verify evidence from code/tests/logs, de-duplicate findings, and produce a severity-ranked report for release gates or comprehensive code review.