openpaper AGENTS.md

Project instructions for OpenPaper, including how Docker sessions write logs into the host's Claude project directory.

In plain words
What is it for?
Use them to locate OpenPaper container logs and identify interactive or test-run sessions by their JSONL metadata and permission settings.
Why use it?
They help distinguish container-created agent sessions from ordinary sessions when investigating logs or testing the Docker setup.

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

Made for: Codex, OpenCode.

Per session 260 This file is loaded in full into every session.
When invoked 260 The same file — it is already loaded in full.
Security scan B 1 finding. 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 $0.00260 $0.00260
Opus 5 $0.00130 $0.00130
Sonnet 5 $0.00052 $0.00052
Haiku 4.5 $0.00026 $0.00026

Measured 2d ago against content hash 2b55de62b40a, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade B, and why

openpaper AGENTS.md scanned grade B 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.

Reads agent configuration directoriesmediumAgent snooping

.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.

grep -l 'bypassPermissions' ~/.claude/projects/-home-sondre-Repositories-OpenPaper/*.jsonl
AGENTS.md · 16 lines

What it actually says

OpenPaper

Docker Session Logs

The docker-compose.yml bind-mounts ~/.claude into the container, so containerized runs (both the interactive service and the test service via test_flow.py) write session logs to the same project directory as host sessions: ~/.claude/projects/-home-sondre-Repositories-OpenPaper/.

To identify container-originated sessions:

  • Agent SDK sessions (test_flow.py dual-agent harness): first JSONL line has "type":"queue-operation" and messages use "permissionMode":"bypassPermissions". The initial user message is the test script's prompt ("I want to set up OpenPaper…").
  • Interactive container sessions (docker-compose --profile interactive): first line has "type":"mode" (like normal CLI sessions) but also use "permissionMode":"bypassPermissions". Cross-reference with ~/.claude/sessions/*.json where the session metadata may include a name field (e.g. "docker-setup-openpaper").

Quick grep to list all container sessions:

grep -l 'bypassPermissions' ~/.claude/projects/-home-sondre-Repositories-OpenPaper/*.jsonl
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 · 16 lines · 260 tokens per session scan B 2b55de62b40a

Subscribe to this mod's changes

openpaper AGENTS.md is an instructions file published in the GitHub repository falense/openpaper (15 stars, last pushed 2mo ago), licensed MIT. It adds 260 tokens to every session, about $0.0013 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.