agentmods.dev

openase: Instructions file for Codex

AGENTS.md

openase AGENTS.md is an instructions file for Codex, OpenCode from PacificStudio/openase. It costs 1,518 tokens per session, scanned A, original, Apache-2.0.

Workspace notes for OpenASE, covering local toolchains and known development issues. They explain how to find Go and Node when the system versions are missing or too old.

In plain words
What is it for?
Use it when running OpenASE’s Go tests, formatting code, checking its web frontend, setting up embedded PostgreSQL tests, or diagnosing startup races.
Why use it?
They provide fallback commands for running tests and formatting, and warn about database and startup problems that can produce misleading failures. This helps agents validate changes in the intended environment.

Instructions file for CodexOpenCode

Written for Codex and OpenCode: the file is AGENTS.md. Also seen: reads .claude/ paths; mentions Claude Code; mentions Codex.

This is PacificStudio/openase's own configuration. It tells Codex and OpenCode how to work on openase itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything openase configures →

not rated 269repo 2mo ago A scan 1,518 tokens original Apache-2.0
View source ↗ PacificStudio/openase pacificstudio.ai
Reuse

Borrowing it

Nothing to install: this file belongs to PacificStudio/openase. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/PacificStudio/openase/main/AGENTS.md
Clone the repo
git clone --depth 1 https://github.com/PacificStudio/openase

Made for: Codex, OpenCode.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for openase AGENTS.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/pacificstudio/openase/agents-md/github.svg)](https://agentmods.dev/instructions/pacificstudio/openase/agents-md)
Your own site
<a href="https://agentmods.dev/instructions/pacificstudio/openase/agents-md"><img src="https://agentmods.dev/badge/instructions/pacificstudio/openase/agents-md/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for openase AGENTS.md

Your own site · 80×15
<a href="https://agentmods.dev/instructions/pacificstudio/openase/agents-md"><img src="https://agentmods.dev/badge/instructions/pacificstudio/openase/agents-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 1,518 This file is loaded in full into every session.
When invoked 1,518 The same file — it is already loaded in full.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.01518 $0.01518
Opus 5.5 $0.00607 $0.00607
Sonnet 5.5 $0.00304 $0.00304
Haiku 4.5 $0.00152 $0.00152

Measured 1mo ago against content hash b56f94e6fb65, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

openase 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 1mo 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 · 24 lines

How it starts

The opening of the file, as written. The whole thing — 24 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Workspace Notes

Environment

  • This workspace may not have go or gofmt on PATH. If Go validation is needed, first try the workspace-local toolchain under .tooling/go/bin/, for example PATH=$PWD/.tooling/go/bin:$PATH go test ./....
  • If .tooling/go/bin is unavailable, a verified fallback toolchain is installed at $HOME/.local/go1.26.1/bin/go; prepend PATH=$HOME/.local/go1.26.1/bin:$PATH for builds, tests, and gofmt.
  • The frontend toolchain may also require a newer Node than /usr/bin/node provides. If Svelte/Vite startup fails on node:util.styleText, use PATH=$HOME/.nvm/versions/node/v22.22.1/bin:$PATH for vitest, prettier, and other web/ checks.
  • In embedded-postgres backed tests, creating ProjectRepo.labels or Agent.capabilities via direct Ent builders currently serializes TEXT[] as JSON-like literals and fails with pq: malformed array literal. Prefer API/service paths that already normalize array input, or avoid asserting on those fields in direct Ent setup until the repository-side array encoding is fixed.
  • Fresh empty databases can currently race schema migration when openase all-in-one starts serve and orchestrate concurrently. For local validation against a brand-new DB, pre-run one migration pass first (for example by opening the DB once through internal/runtime/database.Open) before launching all-in-one.
  • internal/webui/static/.keep must remain tracked on clean checkouts. CI jobs like make openapi-check compile internal/webui/ui.go without first building frontend assets, and deleting the placeholder causes go:embed all:static: no matching files found.
  • Local redeploy cleanup must identify repo-local openase processes by resolved /proc/<pid>/exe, not only by argv. Older runs may survive as <repo-root>/bin/openase (deleted) and keep serving stale code on 127.0.0.1:19836 until explicitly killed.
  • Tests that touch ~/.openase/... must isolate HOME per test run, preferably with t.TempDir() or a test-name-derived temp home. When parallel execution exposes cross-test contamination, fix the isolation boundary first instead of lowering concurrency to hide the problem.
  • Ticket-runtime go test runs can inherit OPENASE_TICKET_ID and similar OpenASE env vars from the harness. Tests that assert missing ticket context should scrub inherited env vars explicitly instead of relying on the parent process env being empty.
  • scripts/ci/backend_coverage.sh rewrites HOME to a temp directory before make check. In this workspace export OPENASE_PGTEST_SHARED_ROOT=$HOME/.cache/openase/pgtest so internal/testutil/pgtest reuses the extracted binaries instead of a cold temp-home cache that may miss the sibling postgres binary next to initdb.
  • Backend CI runs on GitHub-hosted runners. Prefer OPENASE_GO_TEST_PROGRESS_MODE=json for backend CI jobs so go test emits continuous progress instead of relying on sparse wrapper heartbeats.
  • web/playwright.config.ts currently hard-codes 127.0.0.1:4173 with reuseExistingServer: false. When concurrent workspaces already occupy that port, the local CI gate can fail even if the code is correct; in that case, keep the repo config unchanged and run the equivalent Playwright step with a temporary copied config on an alternate localhost port, then restore any generated internal/webui/static artifacts before committing.
  • scripts/dev/test_ephemeral_chat_blackbox.py can fail against real local providers because rate limits or missing provider env vars surface before chat startup. For deterministic local validation in this workspace, prefer python3 scripts/dev/test_ephemeral_chat_blackbox.py --base-url http://127.0.0.1:19836 --provider-mode fake-codex --timeout-seconds 90.
  • This workspace currently has no X11/Wayland display server and no xvfb-run helper installed. Electron GUI E2E should either run on a machine with a display (or an Xvfb-wrapped CI job) or skip explicitly when DISPLAY / WAYLAND_DISPLAY are absent.
  • pnpm in this workspace can ignore install scripts by default. If Electron is installed but its binary payload is missing, run the bundled installer directly (for example node desktop/node_modules/electron/install.js) before attempting Playwright or packaging flows.
  • Claude Code auth can be misleading in this environment: claude auth status may report loggedIn: true while real prompt execution still fails with 401 Invalid authentication credentials, and claude --debug-file ... -p ... shows OAuth refresh attempts returning 400 from https://platform.claude.com/v1/oauth/token. Current Claude builds also store OAuth metadata under ~/.claude/.credentials.json -> claudeAiOauth, with expiresAt as a millisecond Unix timestamp instead of the older top-level RFC3339 fields. Before treating Claude as ready here, verify with a real claude -p ... prompt, not status output alone.
  • Additional Claude auth note: the current CLI also exposes CLAUDE_CODE_OAUTH_TOKEN, but exporting the stale claudeAiOauth.accessToken into that env var still fails real prompts with 401 Invalid authentication credentials, and the stored refreshToken fails with 401 Invalid bearer token. claude auth login and claude setup-token both fall back to interactive browser OAuth flows on this machine, so unattended runs still need an externally refreshed login or a valid ANTHROPIC_API_KEY.
  • Further Claude auth note: this machine's saved Snap Firefox profile at ~/snap/firefox/common/.mozilla/firefox/6sfz7ccx.default/ does contain real claude.ai / claude.com / platform.claude.com cookies, but that still does not unlock unattended Claude Code auth. Raw HTTP reuse of those cookies against Claude OAuth can still hit Cloudflare 403 Just a moment..., and a real headless Firefox session with the saved cookies can get past Cloudflare but still lands on https://platform.claude.com/login?returnTo=%2F%3F with the normal Claude Platform sign-in page instead of an approval/code page. Treat that as more evidence that unattended runs here need a valid ANTHROPIC_API_KEY or a fresh human-completed platform login.
  • Official Claude Code settings docs now also mention ANTHROPIC_AUTH_TOKEN, but on this machine injecting the stored claudeAiOauth.accessToken into ANTHROPIC_AUTH_TOKEN still fails real claude -p --bare ... prompts with 401 and message OAuth authentication is currently not supported. Injecting the stored refreshToken into ANTHROPIC_AUTH_TOKEN fails with 401 Invalid bearer token. So neither documented bearer-token path nor Claude's stored OAuth tokens provide an unattended escape hatch here.

Read the full file on GitHub · 24 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. 1mo ago First seen · 24 lines · 1,518 tokens per session scan A b56f94e6fb65

Subscribe to this mod's changes

openase AGENTS.md is an instructions file published in the GitHub repository PacificStudio/openase (269 stars, last pushed 2mo ago), licensed Apache-2.0. It adds 1,518 tokens to every session, about $0.0061 per session on Opus 5.5. 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.

agentaisoftware-engineering
Related

Other instructions, from other repositories

vscode buildNext.instructions.md

Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).

microsoft/vscode · 9,715 tokens

next.js AGENTS.md

AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.

vercel/next.js · 7,581 tokens

vscode oss-third-party-notices.instructions.md

Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, cache eligibility and rollout and applying the notice (cutover).

microsoft/vscode · 5,262 tokens

langchain AGENTS.md

AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.

langchain-ai/langchain · 4,680 tokens

vscode copilot-instructions.md

Copilot instructions for microsoft/vscode, covering vs code copilot instructions, project overview, root folders, core architecture (src/ folder) and built-in extensions (extensions/ folder).

microsoft/vscode · 2,731 tokens

deepseek-harness AGENTS.md

AGENTS.md instructions for deepseek-ai/deepseek-harness, covering agents.md, pre-stable apis and released session data, repository layout, commands and host sandbox failures.

deepseek-ai/deepseek-harness · 4,261 tokens