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/bitwise-media-group/evolve/agents-mdgit clone --depth 1 https://github.com/bitwise-media-group/evolveWhat 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 | $0.02898 | $0.02898 |
| Opus 5 | $0.01449 | $0.01449 |
| Sonnet 5 | $0.00580 | $0.00580 |
| Haiku 4.5 | $0.00290 | $0.00290 |
Grade A, and why
evolve 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 — 139 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent orientation
Fast map of this repository so a new session can act without re-exploring. For why the code is shaped this way — conventions, the engine/reporter architecture, and how the TUI is wired — read DESIGN.md; for end-user usage read README.md. This file is the "where things are"; DESIGN.md is the "how it fits together".
What this is
evolve is a single Go CLI (module github.com/bitwise-media-group/evolve, entry cmd/evolve) that evaluates
coding-agent plugins across tiers: static checks (Tier 0), trigger-accuracy evals (Tier 1), and behavioral evals (Tier
2), then writes committed Markdown/JSON reports. It drives real agent CLIs (Anthropic, OpenAI, Google, Cursor, Copilot,
Antigravity) in throwaway workspaces and grades the results.
Layout
cmd/evolve/ CLI: package main, one file per verb (cobra). See "Commands" below.
internal/ All library code, one package per concern. See "Packages" below.
docs/ Zensical site (published; nav in zensical.toml). GENERATED by `make docs`:
cli/, man/, and the config/ artifacts (config.schema.json, .evolve.{yaml,jsonc},
.includes/reference.md). The rest — incl. config/index.md — is authored Markdown.
completions/ GENERATED by `make docs`: static bash/zsh/fish completions, shipped in release
archives so the Homebrew cask never executes the binary at install time.
schemas/ JSON Schemas for eval/results/report files (embedded via schemas.go).
e2e/ SEPARATE Go module: live smoke test + fixture repos/ and golden/ outputs.
.mise/ Shared toolchain (bitwise-media-group/toolchain): a git submodule pinned to a
semver tag (.gitmodules `branch`); carries the go-cli task archetype and the mise
pins for the dev CLIs (golangci-lint, goreleaser, addlicense, ...).
security/ Committed code-scanning notes.
.github/workflows/ CI, security (CodeQL), release (goreleaser), merge (fast-forward + auto-merge). Dep
bumps come from the org's Renovate bot (bitwise-media-group/renovate-config).
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 · 139 lines · 2,898 tokens per session scan A 11ee15d26bbe
evolve AGENTS.md is an instructions file published in the GitHub repository bitwise-media-group/evolve (5 stars, last pushed 2d ago), licensed MIT. It adds 2,898 tokens to every session, about $0.0145 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
tokscale AGENTS.md
AGENTS.md instructions for junhoyeo/tokscale, covering ai agent guidelines, maintaining agents.md files, authoring pr/issue content via gh cli, git identity & merge discipline and commit message convention.
codedb AGENTS.md
AGENTS.md instructions for justrach/codedb, covering codedb agent guidelines, what codedb is (and isn't), review guidelines, pre-merge verification and security-sensitive areas.
codedb copilot-instructions.md
Copilot instructions for justrach/codedb, covering codedb — agent instructions, project, rules, filing issues and test style.
codedb CLAUDE.md
Claude Code instructions for justrach/codedb, covering codedb — agent instructions, project, rules, filing issues and test style.
goodboy AGENTS.md
Instructions for akhayam99/goodboy, covering agents.md, where new code goes, naming, components and exports and store selectors and memoization.
goodboy CLAUDE.md
Instructions for akhayam99/goodboy, covering goodboy (claude-specific notes), stack quick-reference and claude-specific notes.