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/tugrulguner/summonpot/agents-mdgit clone --depth 1 https://github.com/tugrulguner/summonpotWrote 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.
[](https://agentmods.dev/instructions/tugrulguner/summonpot/agents-md)<a href="https://agentmods.dev/instructions/tugrulguner/summonpot/agents-md"><img src="https://agentmods.dev/badge/instructions/tugrulguner/summonpot/agents-md.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00858 | $0.00858 |
| Opus 5 | $0.00429 | $0.00429 |
| Sonnet 5 | $0.00172 | $0.00172 |
| Haiku 4.5 | $0.00086 | $0.00086 |
Grade A, and why
summonpot 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 5d 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 — 102 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
Working notes for anyone — human or agent — changing summonpot.
For using summonpot, run summonpot add skills; that installs the endpoint contract
as a skill. This file is about developing the framework itself.
What summonpot is
A contract-first API framework. An endpoint declaration is the executable contract:
request model + docstring goal + declared capabilities + response model
The ellipsis is a complete declaration body, not an unfinished implementation. The decorated declaration rejects direct calls. Today every request runs through the provider-neutral agent runtime; the target is a compiler that picks the least-powerful sufficient executor, skipping the model entirely when one legal path remains.
Two invariants govern almost every design argument:
- The endpoint declaration is the whole contract. Anything that pushes configuration outside the endpoint declaration is moving in the wrong direction.
- No executor may add capabilities, weaken validation, or change the response contract.
Layout
src/summonpot/
summon.py @summon, registration and all registration-time guards
runtime.py the agent loop, model resolution, usage limits and timeout
server.py FastAPI route construction, request schemas, HTTP error mapping
models.py EndpointDef, ToolDef, ParamDef
dependencies.py Depends / Required
tools.py capability construction from callables
_annotations.py shared annotation resolution (used by summon.py and tools.py)
cli.py the summonpot command
commands/ CLI subcommands
skills/ the shipped agent skill, and per-agent formatting
templates/skills/ the skill body itself
Coding guidelines
Rules for changing framework source — including the ones that each cost a real bug —
live in src/summonpot/AGENTS.md. Agents read the nearest
file in the directory tree, so editing source picks them up automatically.
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.
- 5d ago First seen · 102 lines · 858 tokens per session scan A ec8b9759874d
summonpot AGENTS.md is an instructions file published in the GitHub repository tugrulguner/summonpot (2 stars, last pushed today), licensed MIT. It adds 858 tokens to every session, about $0.0043 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
fava-trails AGENTS.md
AGENTS.md instructions for MachineWisdomAI/fava-trails, covering fava trails — agent reference 🫛👣, repository workflow, scope discovery, session start protocol and scope lookup discipline.
deepagents AGENTS.md
AGENTS.md instructions for langchain-ai/deepagents, covering global development guidelines for the deep agents monorepo, corridor security analysis, development workflow, suppressing ruff rules and pr conventions.
mini-swe-agent copilot-instructions.md
Copilot instructions for SWE-agent/mini-swe-agent, covering style guide, test style, bad and good.
valuecell AGENTS.md
Instructions for ValueCell-ai/valuecell, covering guidelines, python programming, python environment, imports and runtime checks.
mda CLAUDE.md
Instructions for sno-ai/mda, covering claude.md — sno-ai/mda, release process for @markdown-ai/cli, to cut a release (any 1.1.x, 1.2.0, 2.0.0 — same flow), release notes — non-negotiable and what's new.
fava-trails CLAUDE.md
Claude Code instructions for MachineWisdomAI/fava-trails, covering fava trails development, where to work, worktree conventions, ci requirements and local development.