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/snchimata/tokenfold/agents-mdgit clone --depth 1 https://github.com/snchimata/tokenfoldWhat 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 | $0.00894 | $0.00894 |
| Opus 5 | $0.00447 | $0.00447 |
| Sonnet 5 | $0.00179 | $0.00179 |
| Haiku 4.5 | $0.00089 | $0.00089 |
Grade A, and why
tokenfold AGENTS.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 2d 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 — 84 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Repository Guidelines
These instructions apply to the entire repository. More specific AGENTS.md files may override them for their directory.
Working principles
- Make the smallest correct change and reuse existing patterns before adding abstractions or dependencies.
- Read the implementation, its callers, and nearby tests before editing behavior.
- Keep changes focused. Do not modify unrelated work already present in the working tree.
- Treat public CLI output, serialized reports, configuration, and language bindings as compatibility-sensitive interfaces.
- Never commit secrets, local configuration, generated build output, or benchmark/evaluation caches.
Repository map
crates/tokenfold-core: compression pipeline, policies, transforms, reports, and token accounting.crates/tokenfold-cli: command-line interface and end-to-end CLI behavior.crates/tokenfold-proxy: HTTP proxy and forwarding behavior.crates/tokenfold-py: PyO3 Python bindings.crates/tokenfold-{adapters,admin,image,learn,output,rag}: optional integration and feature crates.packages/tokenfold: TypeScript API and npm CLI wrapper.tests/: shared fixtures and byte-exact golden cases.python-tests/: tests for the built Python extension.eval/: Python fidelity harness and transform evaluation.- Design notes (architecture, interfaces, testing policy, roadmap decisions) are kept untracked and are not part of a clone; treat the code, tests, and
CHANGELOG.mdas the authoritative record.
Implementation conventions
- Use the pinned Rust toolchain from
rust-toolchain.tomland the workspace's Rust 2024 edition. - Put shared compression behavior in
tokenfold-core; keep CLI, proxy, and bindings as thin adapters where practical. - Preserve JSON insertion order and stable report fields. Do not change schema versions, output formats, exit codes, or defaults accidentally.
- Preserve fail-closed safety behavior, redaction guarantees, protected content, and token-budget accounting.
- Keep lossy or unvalidated transforms behind the existing experimental/fidelity gates.
- Add tests beside the affected behavior. Prefer focused regression tests over broad new test infrastructure.
- Golden outputs are versioned contracts. Change them only for an intentional behavior change, explain why, and keep
tests/golden/MANIFEST.tomlsynchronized. - Avoid new dependencies unless existing workspace crates or the standard library cannot reasonably solve the problem.
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.
- 2d ago First seen · 84 lines · 894 tokens per session scan A bf0c6e4414fa
tokenfold AGENTS.md is an instructions file published in the GitHub repository snchimata/tokenfold (22 stars, last pushed 9d ago), licensed Apache-2.0. It adds 894 tokens to every session, about $0.0045 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 2026-08-30.
Other instructions, from other repositories
headroom copilot-instructions.md
Copilot instructions for headroomlabs-ai/headroom: When performing a pull request review in this repository.
OmniGlyph CLAUDE.md
Instructions for diegosouzapw/OmniGlyph, covering claude.md, quick start, project at a glance, key conventions and running benchmarks.
OmniGlyph AGENTS.md
Instructions for diegosouzapw/OmniGlyph, covering omniglyph — agent guidelines, project, stack, build / test commands and architecture.
OmniGlyph GEMINI.md
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.
distil CLAUDE.md
Claude Code instructions for munhq/distil, covering distil — context optimization for llm agents, what this is, consumption modes, architecture and layers (src/layers/).
densely CLAUDE.md
Instructions for alibaizhanov/densely: Behavioral guidelines to reduce common LLM coding mistakes. Merge with project-specific instructions as needed.