agentmods.dev

charon AGENTS.md

Repository instructions for developing Charon, a Go command-line tool that switches several coding agents between saved endpoints and credential profiles. They focus on protecting real configuration files and credentials during development.

In plain words
What is it for?
For safely developing and testing Charon, including sandboxing the home directory, isolating configuration paths, and avoiding real Keychain access.
Why use it?
They prevent tests and development commands from changing live accounts, settings, or macOS Keychain entries.

Instructions file for CodexOpenCode

11 1mo ago A scan 1,393 tokens original MIT
View the source file mingtheanlay/charon
Install

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.

agentmods
npx agentmods add instructions/mingtheanlay/charon/agents-md
Clone the repo
git clone --depth 1 https://github.com/mingtheanlay/charon

Made for: Codex, OpenCode.

Per session 1,393 This file is loaded in full into every session.
When invoked 1,393 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.01393 $0.01393
Opus 5 $0.00696 $0.00696
Sonnet 5 $0.00279 $0.00279
Haiku 4.5 $0.00139 $0.00139

Measured 2d ago against content hash 52b7fe3d0626, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

charon 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.

AGENTS.md · 112 lines

How it starts

The opening of the file, as written. The whole thing — 112 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md

Guidance for AI coding agents (and humans) working in this repository. charon is a small Go CLI that detects the Codex, Claude Code, OpenCode, and Pi CLIs and switches each one's endpoint + credentials between named profiles.

Golden rule: this tool edits real user credentials

charon reads and writes live config for other tools (~/.codex, ~/.claude, ~/.config/opencode, ~/.local/share/opencode, ~/.pi/agent) and the macOS Keychain. It also reads ~/.claude.json (oauthAccount.emailAddress) solely to name an account-backup profile — that file is never written or snapshotted.

  • Never run charon add, charon switch, charon save, or the interactive menu against your real $HOME while developing. Always sandbox: 
    HOME=$(mktemp -d) go run ./cmd/charon status
  • Tests must never touch real config. Use t.Setenv("HOME", t.TempDir()) and t.Setenv("XDG_CONFIG_HOME", t.TempDir()). See internal/tools/tools_test.go and internal/profile/store_test.go for the pattern.
  • Do not add tests that read or write the real Keychain. The keychain shell-out (internal/secret/keychain_darwin.go) is intentionally left uncovered.
  • Preserve the safety guarantees: atomic writes (temp file + rename), 0600 on credential files / 0700 on dirs, and an auto-backup before every switch. Don't regress these.

Commands

make build      # build ./charon
make test       # go vet + go test -race ./...
make cover      # coverage summary
make lint       # golangci-lint run
make fmt        # gofmt -w .
make run        # build + open the interactive menu (sandbox your HOME first)

Always run make fmt and make test before finishing a change. CI (.github/workflows/ci.yml) runs fmt-check, vet, -race tests, build, and golangci-lint on Linux + macOS; keep all of them green.

Architecture

cmd/charon/         CLI entrypoint (thin; no business logic)
  main.go           main, subcommand dispatch, usage
  commands.go       one cmd* func per subcommand + requireTool
internal/artifact/  snapshot/restore primitives, no tool knowledge
  Artifact/Rotator/Merger/Peeker interfaces; FileArtifact,
  MergedFileArtifact, KeychainArtifact; AtomicWrite
internal/tools/   per-tool adapters
  tool.go           Tool struct, AuthSpec, registry (All/Find)
  providers.go      guards for the shared "charon" provider entry (codex/opencode)
  edit.go           JSON/TOML load-merge-write helpers (preserve unknown keys)
  codex.go / claude.go / opencode.go / pi.go   one file per tool
internal/profile/ snapshot store, split by concern:
  store.go (layout/config/name validation) · snapshot.go (Save/Add/Edit/EnsureDefault)
  apply.go (Apply/Undo/Drift/refresh) · backup.go (backups + prune) · manage.go (rm/mv/cp)
internal/models/  fetch model lists from a provider API (openai/anthropic wire)
internal/secret/  masking + platform keychain (darwin vs. other build tags)
internal/tui/     bubbletea interactive menu

Read the full file on GitHub · 112 lines

Changes

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.

  1. 2d ago First seen · 112 lines · 1,393 tokens per session scan A 52b7fe3d0626

Subscribe to this mod's changes

charon AGENTS.md is an instructions file published in the GitHub repository mingtheanlay/charon (11 stars, last pushed 1mo ago), licensed MIT. It adds 1,393 tokens to every session, about $0.0070 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.

9routerai-agentsbyokclicliproxyapi
Related

Other instructions, from other repositories

open-design AGENTS.md

AGENTS.md instructions for nexu-io/open-design, covering directory guide, core documentation index, workspace directories, inactive or placeholder directories and development workflow.

nexu-io/open-design · 10,311 tokens

open-codesign CLAUDE.md

Instructions for OpenCoworkAI/open-codesign, covering claude.md — open codesign, what this project is, hard constraints (do not violate), stack & conventions and frontend stack (locked).

OpenCoworkAI/open-codesign · 1,789 tokens

html-anything AGENTS.md

Instructions for nexu-io/html-anything, covering this is not the next.js you know, workspace shape and commands.

nexu-io/html-anything · 386 tokens

k-dense-byok AGENTS.md

Instructions for K-Dense-AI/k-dense-byok, covering claude.md, project overview, commands, architecture: how a turn flows and project / sandbox layout.

K-Dense-AI/k-dense-byok · 7,467 tokens

nexu AGENTS.md

Instructions for nexu-io/nexu, covering agents.md, repo overview, project overview, commands and branch model.

nexu-io/nexu · 7,138 tokens

open-codesign AGENTS.md

Instructions for OpenCoworkAI/open-codesign, covering agents.md - open codesign, what this project is, hard constraints, ai visibility for web work and current architecture direction.

OpenCoworkAI/open-codesign · 3,343 tokens