stetra AGENTS.md

Repository-specific instructions for Stetra, a system that connects developer intent, coding-agent work, runtime facts, a structured handoff, and a human adoption decision.

In plain words
What is it for?
They are for making changes to Stetra's positioning, lifecycle, authority, persistence, public interfaces, host interaction, command-line protocol, and task behavior.
Why use it?
They explain what the product is and is not, where its architecture is documented, and which boundaries must be respected before changing code or behavior.

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/sovea/stetra/agents-md
Clone the repo
git clone --depth 1 https://github.com/Sovea/stetra

Made for: Codex, OpenCode.

Per session 3,400 This file is loaded in full into every session.
When invoked 3,400 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 $0.03400 $0.03400
Opus 5 $0.01700 $0.01700
Sonnet 5 $0.00680 $0.00680
Haiku 4.5 $0.00340 $0.00340

Measured 2d ago against content hash 44a718d63631, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

stetra 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 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.

AGENTS.md · 368 lines

How it starts

The opening of the file, as written. The whole thing — 368 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md

What this repository builds

Stetra is a cognition-preserving adaptive delegation harness for production coding. It connects exact developer intent, Agent delivery, Runtime-collected facts, a decision-oriented Cognitive Handoff, and the Human adoption decision without weakening developer understanding or engineering judgment.

It is not a coding agent, automated approver, generic project manager, prompt library, repository wiki, transcript store, broad memory system, worktree fleet, or universal workflow engine.

Read docs/architecture.md before changing product positioning, lifecycle, authority, persistence, public APIs, Host interaction, or long-term boundaries. That document contains both implemented and planned architecture; planned capabilities are not implementation authorization. Read docs/change-workflow.md before changing the current CLI protocol or task behavior. Always distinguish current implementation, target architecture, and measured product evidence.

Product architecture

The target design is two planes, three logical graphs, and one append-only ledger:

  • Adaptive Delivery Plane — Agent-owned investigation, delivery strategy, bounded loops, and eventually justified conditional or parallel execution.
  • Deterministic Assurance and Decision Plane — authority, contracts, facts, evidence ceilings, handoff, and Human decision boundaries.
  • Delivery Graph — task-local Agent strategy; planned beyond the current fixed lifecycle.
  • Evidence Graph — task- and Attempt-scoped provenance from Human Event to evidence, conclusion, and review surface; its core relationships exist now.
  • Decision Graph — longitudinal Human-authorized decisions and outcomes; not implemented.
  • Append-only ledger — immutable task artifacts and rebuildable projections; implemented only at current-task scope.

The three task cores are:

  1. Semantic Contract — what one change is intended and authorized to mean.
  2. Fact Spine — what the workflow actually observed.
  3. Cognitive Handoff — what the change means, what remains unknown, and where direct review has adoption value.

Read the full file on GitHub · 368 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. 2d ago First seen · 368 lines · 3,400 tokens per session scan A 44a718d63631

Subscribe to this mod's changes

stetra AGENTS.md is an instructions file published in the GitHub repository Sovea/stetra (2 stars, last pushed 2d ago), licensed MIT. It adds 3,400 tokens to every session, about $0.0170 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.