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 agentmods add instructions/tyejcoleman/tokenroom/claude-mdgit clone --depth 1 https://github.com/tyejcoleman/tokenroomWrote 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/instructions/tyejcoleman/tokenroom/claude-md)<a href="https://agentmods.dev/instructions/tyejcoleman/tokenroom/claude-md"><img src="https://agentmods.dev/badge/instructions/tyejcoleman/tokenroom/claude-md.svg" alt="Measured on agentmods" 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 | $0.01710 | $0.01710 |
| Opus 5 | $0.00855 | $0.00855 |
| Sonnet 5 | $0.00342 | $0.00342 |
| Haiku 4.5 | $0.00171 | $0.00171 |
Grade A, and why
tokenroom CLAUDE.md 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 5d 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 — 108 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Tokenroom — agent working context
Tokenroom is a resource-aware layer for coding agents. It makes the agent aware of its two scarcest resources — account rate-limit headroom (5h / 7d windows) and session context headroom (tokens before compaction) — so it can plan work to fit, spend efficiently, and use capacity that would otherwise expire.
Read this before any work
docs/ARCHITECTURE.md— what talks to what: data flow, module map, extension points.docs/DECISIONS.md— the ADR log: every standing decision with its why. Do not silently violate an ADR; propose a new one instead.docs/PLAN.md— the phased plan; every task has acceptance criteria (AC).- Background when needed:
docs/ONE-PAGER.md(original design),docs/VALIDATION.md(assumption gates + eval methodology).
Hard gates (enforced, not advisory)
scripts/check-invariants.mjs runs after every Edit/Write you make (repo PostToolUse
hook) and inside npm test/CI. It blocks: new dependencies or non-builtin imports
(ADR-2), any network surface (ADR-1/G3), compliance tripwires (ADR-1/G4), and
catch-free tap/hook entry points (ADR-5). If a gate blocks you and you believe you're
right, stop and propose an ADR change — never work around a gate.
Procedures (repo slash-commands)
/release (version → CHANGELOG → tag → publish checklist) · /add-fixture (new payload
shape → sanitized fixture + tests) · /run-evals (required for any stamp/skill/verdict
wording change, per ADR-9).
Hard rules (compliance — non-negotiable)
These are the project's reason to exist responsibly. Any change that touches them is wrong:
- Official extension points only: statusline stdin JSON, hooks (
UserPromptSubmit,SessionStart,PreCompact,PreToolUse), MCP, OTel, and the user's own~/.claude/~/.codexfiles. - Never reuse a subscription OAuth token outside the official client, call
undocumented endpoints (e.g.
api/oauth/usage), spoof harness identity headers, read.credentials.jsonto make API calls, or burn interactive subscription quota headlessly. (The ADR-16 carve-out that once allowed user-armed headless spend is withdrawn — ADR-22 removed the ARM executor; autonomous continuation belongs to the separate Conductor package on in-session surfaces.) - Defensive parsing always: every field we read may be absent, malformed, or buggy
(e.g. epoch values leaking into
used_percentage). Degrade gracefully, clamp to valid ranges, never crash the statusline.
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.
- 5d ago First seen · 108 lines · 1,710 tokens per session scan A 7ce6334e74a2
tokenroom CLAUDE.md is an instructions file published in the GitHub repository tyejcoleman/tokenroom (0 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 1,710 tokens to every session, about $0.0086 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 instructions, from other repositories
Accordion CLAUDE.md
Instructions for a-Fig/Accordion, covering claude.md — accordion, key urls, terminology, codebase map and engine — single source of truth.
tokf CLAUDE.md
Claude Code instructions for mpecan/tokf, covering tokf — development guidelines, project philosophy, commits, code quality and testing.
Accordion AGENTS.md
Instructions for a-Fig/Accordion, a project described as: 🏆 AI Hackathon 2026 @ UC Berkeley Intelligent context management for developers.
mentedb copilot-instructions.md
Copilot instructions for nambok/mentedb, covering mentedb development instructions, project overview, workspace structure, build, test, and lint and key types.
OmniGlyph CLAUDE.md
Claude Code instructions for diegosouzapw/OmniGlyph, covering claude.md, quick start, project at a glance, key conventions and running benchmarks.
OmniGlyph GEMINI.md
Gemini CLI instructions for diegosouzapw/OmniGlyph, covering omniglyph — guidelines for gemini-based agents, non-negotiables (mirror of agents.md), file placement and validation before you claim done.