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/dyingg/chromectx/agents-mdgit clone --depth 1 https://github.com/dyingg/chromectxWhat 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 | $0.01866 | $0.01866 |
| Opus 5 | $0.00933 | $0.00933 |
| Sonnet 5 | $0.00373 | $0.00373 |
| Haiku 4.5 | $0.00187 | $0.00187 |
Grade A, and why
chromectx AGENTS.md scanned grade A with 1 finding 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.
Runs shell commandslowCapability
Expected in a hook, worth knowing in a rule or an instructions file.
- Source code must use `node:` stdlib modules (`node:child_process`, `node:fs/promises`, `node:path`, etc.) for portability. Do **not** use Bun-specific APIs (`Bun.spawn`, `Bun.write`, `Bun.file`, `Bun.Glob`) in producti How it starts
The opening of the file, as written. The whole thing — 97 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Micro-commits
Commit immediately after completing each logical unit of work — do not batch unrelated changes and do not wait to be asked. A "logical unit" is one self-contained change to the codebase: implementing a function, adding a command, writing a test, fixing a bug, wiring up a new module, adding a skill, etc. A unit may touch multiple files if they belong together (e.g. a module + its test, or a skill's directory of files). When in doubt, commit more often rather than less.
Keep commits atomic: commit only the files you touched and list each path explicitly. For tracked files run git commit -m "<scoped message>" -- path/to/file1 path/to/file2. For brand-new files, use the one-liner git restore --staged :/ && git add "path/to/file1" "path/to/file2" && git commit -m "<scoped message>" -- path/to/file1 path/to/file2
Project shape
This repo is a TypeScript codebase (runs on Bun) with one executable surface:
chromectx doctorchromectx restore ...chromectx save ...chromectx list ...chromectx search ...— fast title+URL search;--deepfetches full page contentchromectx rag ...— RAG search across tabs (JSON output for agents/scripts)chromectx mcp
The MCP server is local-only and uses stdio. Treat the CLI and MCP server as two transports around shared code rather than as separate applications.
Source layout
src/bin: executable entrypoints onlysrc/browser: browser-level type definitions (ChromeSession,ChromeTab, etc.) shared across commands, store, and platformsrc/commands: CLI-facing orchestration — each command is a self-contained module exporting aCommandDefinition(see "Adding a CLI command" below);types.tsdefines the sharedCommandDefinition,CommandRunContext, andGlobalFlagsinterfaces;root.tsis the registry + dispatcher — each command is a self-contained module exporting aCommandDefinition(see "Adding a CLI command" below);types.tsdefines the sharedCommandDefinition,CommandRunContext, andGlobalFlagsinterfaces;root.tsis the registry + dispatchersrc/mcp: MCP protocol handling and tool registrationsrc/platform/macos: macOS-specific integrations and checkssrc/platform/macos/chrome/: Chrome browser interaction (install detection, session/tab queries, page source retrieval)src/lib: shared support modules such as config, errors, output, logging,concurrent.ts(sliding-windowpMap), andhttp.ts(sharedfetchSourceswith concurrency 15)src/lib/tui/: interactive terminal UI components built on@clack/prompts—select.tsexportsselectOne<T>()(and will gainselectMany<T>()later). Use these for any arrow-key selection in CLI commands; inject viadepsfor testing.src/lib/store/: session persistence —types.tsdefines the schema contract (Session,SessionWindow,SessionTab);io.tshandles read/write/list to{AppPaths.sessions}/*.jsonsrc/lib/workflows/search/: BM25 search pipeline —preprocess.ts(HTML→markdown via turndown),chunk.ts(paragraph chunking ~500 chars),index.ts(buildIndex+SearchIndexwithsearchandsearchWithPages)src/lib/workflows/rag.ts: shared RAG pipeline (executeRag) used by both CLIragcommand and MCPrag_searchtool
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 · 97 lines · 1,866 tokens per session scan A f8afc9687eb1
chromectx AGENTS.md is an instructions file published in the GitHub repository dyingg/chromectx (4 stars, last pushed 5mo ago), licensed MIT. It adds 1,866 tokens to every session, about $0.0093 per session on Opus 5. A static security scan graded it A with 1 finding (runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
chrome-console-log-mcp-server AGENTS.md
AGENTS.md instructions for ONLYHUMN/chrome-console-log-mcp-server, a project described as: Capture browser console logs via the Chrome DevTools Protocol and expose them to an MCP client over SSE.
flyto-core AGENTS.md
Instructions for flytohub/flyto-core, covering flyto-core agent rules, flyto2 project memory contract, flyto2 frontend quality gate and repo notes.
rosetta pr-review-bots.instructions.md
How Copilot and CodeRabbit reviews manifest on a rosetta PR, why each looks different from a gh point of view, and what 'wrapping a PR' actually requires.
mcp-screenshot-server CLAUDE.md
Instructions for sethbang/mcp-screenshot-server, covering claude.md, what this is, commands, environment variables and architecture.
one-search-mcp AGENTS.md
Instructions for yokingma/one-search-mcp, covering agent guide, start here, project map, implementation rules and tests and verification.
rosetta AGENTS.md
Instructions for tikoci/rosetta, covering codex instructions for rosetta, first reads, development defaults and mcp and client config.