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/tryghost/ghost/agents-mdgit clone --depth 1 https://github.com/TryGhost/GhostWhat 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 | $0.00761 | $0.00761 |
| Opus 5 | $0.00380 | $0.00380 |
| Sonnet 5 | $0.00152 | $0.00152 |
| Haiku 4.5 | $0.00076 | $0.00076 |
Grade A, and why
Ghost 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 — 67 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
Agent-specific execution guidance for the Ghost monorepo. Human-readable setup, workflow, architecture, and practice guidance lives in the codebase documentation and nearby package READMEs.
Start with:
- Development setup
- Contribution workflow
- Writing codebase documentation
- Testing
- Shipping
- Monorepo structure
Required workflow
- Always use
pnpm, never npm or Yarn. External dependency versions belong in the catalogs inpnpm-workspace.yaml; workspace dependencies useworkspace:versions. - Run
pnpm setupbefore other commands in a fresh checkout or worktree. - Use
pnpm checkas the default full validation command. Browser E2E and Ember Admin tests run separately; follow the testing guide. - Read the nearest
AGENTS.md,CLAUDE.md, and README before changing a package or subsystem. More specific guidance overrides this file. - When committing, load and follow
.agents/skills/commit/SKILL.md.
Repository skills
Repository skills live under .agents/skills/. When adding one, also add the
matching .claude/skills/<name> symlink to
../../.agents/skills/<name>. Run pnpm lint:agent-skills to verify discovery.
Use the relevant repository skill before adding an Admin API endpoint, database migration, private feature flag, Shade component, or internal package.
Task routing and important warnings
- Admin UI: read
apps/admin/README.mdandapps/shade/AGENTS.md. Build new features in React, useadmin-x-frameworkfor APIs, and use Shade for UI. Admin and Core deploy independently, so feature-detect backend support and test the older-backend case. - Embedded Admin CSS: do not import
@tryghost/shade/styles.cssfrom an embedded app. Admin owns the single Tailwind and Shade CSS lane. - Translations: follow the
internationalization guide. Run the
extraction command after changing
t()calls and never split one sentence across translation calls. - Public apps: read the app's README and the shipping guide. Their release and CSS lanes differ from Admin.
- Ghost Core: use the server map
and read the services guide before
adding a service. New standalone services use TypeScript; keep CommonJS only
at existing
require()boundaries. Boot owns service initialization; do not initialize on the first request. - ESLint: use the shared factories and dependency rules in the ESLint configuration README. A hand-written config must declare every plugin it imports locally.
- Analytics: start with
pnpm dev:analyticsand follow the nearby Tinybird READMEs underghost/core/core/server/data/tinybird/.
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 · 67 lines · 761 tokens per session scan A 3414ce388319
Ghost AGENTS.md is an instructions file published in the GitHub repository TryGhost/Ghost (55,095 stars, last pushed today), licensed MIT. It adds 761 tokens to every session, about $0.0038 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
moon AGENTS.md
Instructions for moonrepo/moon, covering agents.md, repository layout, prerequisites, rust workflow and conventions.
sonicjs AGENTS.md
Instructions for SonicJs-Org/sonicjs, covering sonicjs agent guidelines, decision tree — code exploration, rtk (rust token killer), caveman mode and delegation — cavecrew subagents.
intlayer CLAUDE.md
Instructions for aymericzip/intlayer, covering repo layout, commands (run from root), content declaration flow, git and commit conventions.
featurevisor AGENTS.md
Instructions for featurevisor/featurevisor, covering links, stack, packages, examples and installation.
bun AGENTS.md
AGENTS.md instructions for oven-sh/bun, a project described as: Incredibly fast JavaScript runtime, bundler, test runner, and package manager – all in one.
bun CLAUDE.md
Claude Code instructions for oven-sh/bun, covering building and running bun, build commands, changes that don't require a build, testing and running tests.