semley AGENTS.md

Repository instructions for Semley, an autonomous site-reliability investigation agent that investigates system problems through a controlled sequence of steps. They explain setup, checks, demos, and how to run it with an OpenAI-compatible model.

In plain words
What is it for?
Use them to run deterministic checks, view a no-key demonstration, configure a model, and start host investigations.
Why use it?
They provide a quick, project-specific path for verifying and running Semley without guessing its prerequisites or workflow.

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

Made for: Codex, OpenCode.

Per session 1,044 This file is loaded in full into every session.
When invoked 1,044 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 $0.01044 $0.01044
Opus 5 $0.00522 $0.00522
Sonnet 5 $0.00209 $0.00209
Haiku 4.5 $0.00104 $0.00104

Measured yesterday against content hash 073441b81945, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

semley 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 yesterday.

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 · 93 lines

How it starts

The opening of the file, as written. The whole thing — 93 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md

Guidance for coding agents (and people) working in this repo. Semley is an autonomous SRE investigation agent: a language model drives an investigation through one governed step tool, a Burr state machine gates which transitions are legal, and conclusions are verified against real evidence in code the model cannot reach. README.md has the full design; this file is the fast path to running it.

Setup

uv sync                                   # needs uv and Python 3.12+
printf 'OPENAI_API_KEY=sk-...\n' > .env   # any OpenAI-compatible key; default model gpt-5.4

The agent is vendor-agnostic (it speaks the OpenAI chat-completions protocol). For a non-OpenAI vendor, also set OPENAI_BASE_URL (e.g. https://openrouter.ai/api/v1) and SEMLEY_MODEL (e.g. anthropic/claude-sonnet-5).

Verify it works (no key, no infrastructure)

make check                                # 7 fast deterministic tests; no model, no infra

To see a real end-to-end investigation without a key or any infrastructure, read recordings/host-investigation.txt.

Run the agent

uv run semley --surface host              # needs a key; the host demo also needs OrbStack

Describe an incident in plain language, confirm the proposed target and scope, and it drives the investigation to a grounded conclusion. /playbook saves the recorded Ansible reads as a standard playbook, /quit exits, Tab completes the commands.

Demo targets

Each command below stands up its own target (printing clear status) and then runs the agent. These create real local infrastructure, so run them deliberately, not as a quick check:

Command What it does Requires
make demo-host inject a service fault on web1, investigate, restore OrbStack
make demo-cluster kind cluster + a workload in ImagePullBackOff, investigate Docker, kind, kubectl
make demo-telemetry Prometheus + a down scrape target, investigate Docker, kind, kubectl
make demo-localhost investigate this machine nothing extra

Read the full file on GitHub · 93 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. yesterday First seen · 93 lines · 1,044 tokens per session scan A 073441b81945

Subscribe to this mod's changes

semley AGENTS.md is an instructions file published in the GitHub repository msradam/semley (0 stars, last pushed 1mo ago), licensed MIT. It adds 1,044 tokens to every session, about $0.0052 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 2026-09-01.