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/calesthio/generative-media-skills/agents-mdgit clone --depth 1 https://github.com/calesthio/generative-media-skillsWhat 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 | $0.00619 | $0.00619 |
| Opus 5 | $0.00309 | $0.00309 |
| Sonnet 5 | $0.00124 | $0.00124 |
| Haiku 4.5 | $0.00062 | $0.00062 |
Grade A, and why
generative-media-skills 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.
How it starts
The opening of the file, as written. The whole thing — 45 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent instructions
Read METHODOLOGY.md completely before creating, editing, or reviewing a skill.
Non-negotiable rules
- A leaf skill directory at
skills/<family>/<category>/<skill-name>/must containSKILL.mdand hiddenEVAL.md. It may also contain only these optional top-level directories:scripts/,references/,assets/, andtests/. - Category directories contain only subdirectories. Do not add arbitrary leaf-level files, templates, notes, manifests, or top-level directories beyond the allowed package contract.
- Publication includes only
SKILL.mdplusscripts/,references/, andassets/when present.EVAL.mdandtests/are repository-only and must be excluded from bundled skill packages by default. - Root
tools/may contain repository maintenance tooling such as package export; copied skills must not silently depend on root tooling at runtime. - Do not read or imitate sibling skills while authoring a new skill unless the task is explicitly an audit or comparison.
- Do not use another completed skill as a structural, stylistic, or conceptual template.
- Research the assigned provider, model, craft, or production category independently.
- Prefer official documentation, model cards, technical reports, standards, peer-reviewed research, and reproducible tests.
- Label documented facts, empirical observations, and production heuristics distinctly.
- Cite consequential factual claims and record when volatile facts were verified.
- Include complete examples when examples materially improve an agent's ability to produce content. Label every example as an example, not a mandatory formula.
- Use supporting resources for progressive disclosure: keep
SKILL.mdas the navigable entrypoint, reference bundled files with relative paths, and state when an agent should load or run them. - Add scripts only when executable code is materially better than prose. Scripts must be portable, dependency-explicit, safe by default, dry-run capable for side effects, and covered by repository tests.
- Do not impose an arbitrary length limit. Include the depth required by the subject and exclude unrelated material.
- Write for an AI agent performing real content production, not for a human browsing a list of prompt tips.
- Never expose
EVAL.mdto the agent being evaluated. Evaluate with the published package only, then score withEVAL.md.
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.
- yesterday First seen · 45 lines · 619 tokens per session scan A 8ff5d17b6e08
generative-media-skills AGENTS.md is an instructions file published in the GitHub repository calesthio/generative-media-skills (143 stars, last pushed 1mo ago), licensed MIT. It adds 619 tokens to every session, about $0.0031 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 2026-08-30.
Other instructions, from other repositories
OpenStudio CLAUDE.md
Claude Code instructions for palcocerhurtado-tech/OpenStudio, covering openstudio — claude code instructions, your role, key files to read first, bridge tools (new in openstudio) and generate an image.
mini-swe-agent copilot-instructions.md
Instructions for SWE-agent/mini-swe-agent, covering style guide, test style, bad and good.
intelligent-terminal rust.instructions.md
Concise Rust coding conventions for this repository.
AICOO-Skills AGENTS.md
Instructions for Aicoo-Team/AICOO-Skills, covering agents.md, if you just installed this pack (read first), repo purpose, editing rules and documentation sync.
MassGen AGENTS.md
AGENTS.md instructions for massgen/MassGen, covering claude.md, openspec instructions, planning, instruction file parity and debugging assumptions.
MassGen CLAUDE.md
Claude Code instructions for massgen/MassGen, covering claude.md, openspec instructions, planning, instruction file parity and debugging assumptions.