agentic AGENTS.md

Repository instructions for Agentic, a library that stores reusable coding-agent instructions, skills, vendor settings, and project profiles.

In plain words
What is it for?
Use them when editing agent instruction fragments, skills, adapters, profiles, schemas, or the library’s tooling.
Why use it?
They give contributors consistent rules for structuring, writing, and validating those reusable pieces.

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

Made for: Codex, OpenCode.

Per session 1,585 This file is loaded in full into every session.
When invoked 1,585 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 $0.01585 $0.01585
Opus 5 $0.00792 $0.00792
Sonnet 5 $0.00317 $0.00317
Haiku 4.5 $0.00159 $0.00159

Measured 2d ago against content hash ba90c6ddcc17, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

agentic 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 · 125 lines

How it starts

The opening of the file, as written. The whole thing — 125 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Agentic Library — Agent Instructions

This repository is a centralized library of vendor-agnostic agentic configurations. It is never used directly by end-user agents. Its fragments are composed into target project repositories via the tooling provided here.

What This Repo Contains

  • agents/ — Composable AGENTS.md building blocks (fragments)
  • skills/ — Reusable agent skill definitions
  • vendors/ — Vendor-specific adapter configurations
  • profiles/ — Composition presets for common project types
  • tooling/ — Scripts and utilities (called via just)
  • schemas/ — JSON Schemas for all structured files

Rules for Working in This Repo

File Authoring

  • Fragment files in agents/ must be self-contained: no cross-references to other fragments, no project-specific paths, no hardcoded credentials.
  • Every fragment starts with a single H2 heading (## Section Name) that matches its purpose.
  • Fragments must not exceed 300 lines. Split larger content into multiple focused fragments.
  • Use {TOKEN_NAME} syntax for project-specific values that get substituted at compose time (e.g., {BUILD_CMD}, {PROJECT_NAME}).

Skill Authoring

  • Every skill directory must have a SKILL.md with valid YAML frontmatter.
  • Frontmatter must include: name, description, version, tags, vendor_support.
  • Skills must declare all supporting files in resources: frontmatter.
  • Use SemVer for version (MAJOR.MINOR.PATCH).

Vendor Adapters

  • Never edit vendor-generated output files manually. They are derived artifacts.
  • The canonical source of truth is always AGENTS.md (assembled from fragments).
  • adapter.json files must validate against vendors/_schema/adapter.schema.json.

Documentation

  • Any change that introduces, renames, or removes a vendor output file, symlink path, tooling flag, or just recipe must include a corresponding update to the affected docs in the same commit. Relevant files: README.md, docs/vendors.md, docs/commands.md, docs/extending.md, and any vendors/*/README.md.
  • When adding a new vendor adapter, add it to the Supported Vendors table in README.md and docs/vendors.md, and document its output files and symlink paths.
  • When adding a new just recipe or CLI command, add it to docs/commands.md.
  • When adding or updating a skill that references tooling paths or output filenames, verify those paths are still accurate before committing.
  • When skills are added, renamed, regrouped, removed, or their user-visible metadata (for example name/description in skill frontmatter) changes under skills/, update docs/skills.md in the same change.

Read the full file on GitHub · 125 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 · 125 lines · 1,585 tokens per session scan A ba90c6ddcc17

Subscribe to this mod's changes

agentic AGENTS.md is an instructions file published in the GitHub repository soulcodex/agentic (10 stars, last pushed 2d ago), licensed MIT. It adds 1,585 tokens to every session, about $0.0079 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.