coco AGENTS.md

Repository instructions for coco, a Telegram-connected application that keeps a limited log of chat messages for working memory. They document session startup, memory storage, runtime settings, and safe self-updates.

In plain words
What is it for?
Use them when loading chat memory, changing the Telegram memory writer, configuring its log path, or implementing update checks.
Why use it?
They help an agent restore recent Telegram context while protecting user files and avoiding unsafe cleanup during updates.

Instructions file for CodexOpenCode

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/pcopu/coco/agents-md
Clone the repo
git clone --depth 1 https://github.com/pcopu/coco

Made for: Codex, OpenCode.

Per session 351 This file is loaded in full into every session.
When invoked 351 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 $0.00351 $0.00351
Opus 5 $0.00176 $0.00176
Sonnet 5 $0.00070 $0.00070
Haiku 4.5 $0.00035 $0.00035

Measured 2d ago against content hash 64980f009588, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

coco 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 · 35 lines

What it actually says

Repository Memory Notes

Auto Memory Log

  • Memory log path: TELEGRAM_CHAT_MEMORY.jsonl (repo root).
  • Purpose: store only Telegram-visible chat text (incoming message text/captions, outgoing sends, outgoing edits).
  • This is intentionally not a full internal execution transcript.

Session Start Prompt

Use this at the start of a new agent session in this repo:

Load rolling Telegram memory from TELEGRAM_CHAT_MEMORY.jsonl.
Read the latest 200 lines and build a concise working-memory summary:
- active goals
- recent decisions/constraints
- unresolved tasks
- latest verified environment facts
Treat this log as source-of-truth for what was actually shown in Telegram chat.

Runtime Notes

  • Auto-writer is implemented in src/coco/telegram_memory.py.
  • To override log location, set COCO_TELEGRAM_MEMORY_LOG_PATH.

Self-Update Safety

  • CoCo self-update must treat only tracked or staged changes as a dirty worktree. Use git status --porcelain --untracked-files=no for this check.
  • Untracked research, temporary files, and backups are user-owned artifacts. They must not block /update, and update code must never delete, stash, or clean them.
  • Keep the local updater in src/coco/bot.py and the remote-node updater in src/coco/agent_rpc.py aligned when changing update safety behavior.
  • Preserve regression coverage proving that untracked-only worktrees can update while modified tracked files still block. git pull --ff-only must remain the final guard against an incoming tracked path overwriting an untracked file.
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 · 35 lines · 351 tokens per session scan A 64980f009588

Subscribe to this mod's changes

coco AGENTS.md is an instructions file published in the GitHub repository pcopu/coco (4 stars, last pushed 8d ago), licensed MIT. It adds 351 tokens to every session, about $0.0018 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.