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/randomm/pi-ensemble/agents-mdgit clone --depth 1 https://github.com/randomm/pi-ensembleWhat 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 | $0.21658 | $0.21658 |
| Opus 5 | $0.10829 | $0.10829 |
| Sonnet 5 | $0.04332 | $0.04332 |
| Haiku 4.5 | $0.02166 | $0.02166 |
Grade A, and why
pi-ensemble AGENTS.md scanned grade A with 0 findings against 26 rules in 11 categories โ prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency โ measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing โ 631 lines โ stays where its author put it; the contents beside it link to each section on GitHub.
Agent Guidelines for pi-ensemble
This document is auto-loaded by Pi when any agent works inside this repo. It establishes the rules, conventions, and project-specific constraints all development work follows. Read once; apply throughout.
โน๏ธ Recursive context. pi-ensemble both uses Pi (when you, the agent, run inside it) and configures Pi (its actual purpose โ building role prompts and an orchestrator extension). When this file says "the agent", it means you, the one reading this. When it says "subagents" / "role prompts" / "specialists", it means pi-ensemble's product surface.
1. ๐จ Pre-Push Quality Gates (BLOCKING)
CI is for VERIFICATION, not DISCOVERY. All gates pass locally before git push. Never push to "see if CI catches anything."
# 1. Build prompts (repo root) โ generates dist/prompts/standard/*.md
bun run build
# 2. Type-check, lint, and smoke-test (extension/)
cd extension && bunx tsc --noEmit && bun run check && \
for t in smoke-tests/test-*.ts; do \
case "$t" in *-live.ts) continue;; esac; \
bun run "$t" || exit 1; \
done
If any check fails: fix the issue, re-run, only then push. Do NOT:
- โ Use
--no-verify,--no-gpg-sign, or skip hooks - โ Add
// biome-ignoreor// @ts-ignoreto bypass โ fix the code - โ Push and hope CI catches it
- โ Run only the smoke test that exercises your change โ they're all fast offline
Live smoke tests are NOT in the pre-push set
Files matching extension/smoke-tests/test-*-live.ts spawn real Pi children and cost real tokens (a few cents per run on Cerebras GLM-4.7). Run them only when:
- You changed
spawn.ts(changes to child process behaviour) - You bumped the Pi version pin in
extension/package.json - You're investigating a production-only failure mode
CI runs the offline tests only. Live tests are dev-machine only.
Six-pass review findings are blocking at MEDIUM severity and above. The fix loop continues until all MEDIUM, HIGH, and CRITICAL findings are resolved. Only LOW findings may be deferred or overridden with user confirmation.
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 ยท 631 lines ยท 21,658 tokens per session scan A f456ddbfa3ed
pi-ensemble AGENTS.md is an instructions file published in the GitHub repository randomm/pi-ensemble (5 stars, last pushed 2d ago), licensed Apache-2.0. It adds 21,658 tokens to every session, about $0.1083 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.
Other instructions, from other repositories
spec-kit AGENTS.md
AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart โ add a new integration in 5 steps, integration architecture and integrationmanifest โ file tracking.
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
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).
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.
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).
next.js AGENTS.md
Instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.