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/sylin-org/ghostlight/agents-mdgit clone --depth 1 https://github.com/sylin-org/ghostlightWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/sylin-org/ghostlight/agents-md)<a href="https://agentmods.dev/instructions/sylin-org/ghostlight/agents-md"><img src="https://agentmods.dev/badge/instructions/sylin-org/ghostlight/agents-md.svg" alt="Measured on agentmods" height="20"></a>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.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.03293 | $0.03293 |
| Opus 5 | $0.01647 | $0.01647 |
| Sonnet 5 | $0.00659 | $0.00659 |
| Haiku 4.5 | $0.00329 | $0.00329 |
Grade A, and why
ghostlight 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 4d 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.
How it starts
The opening of the file, as written. The whole thing — 231 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Ghostlight -- agent guide
This is the canonical onboarding document for any coding agent (or human) working in this
repository, regardless of which tool or model you are. CLAUDE.md is a thin pointer here.
Assume you have no memory of prior sessions: everything you need is in this file and the
documents it points to.
Read this first (in order)
- docs/MEMORY.md -- the cross-agent project memory: the owner's standing
working preferences, durable cross-cutting learnings, and a pointer index to everything below.
Read it first; it tells you where each kind of durable fact lives. Its sensitive/machine-local
counterpart is
local/NOTES.md(gitignored). - docs/STATUS.md -- where the project stands right now: version state, open PRs, in-flight work, and the owed-items list. Read it before starting anything.
- docs/1.0/INTENT.md, docs/1.0/LANGUAGE.md, docs/1.0/ARCHITECTURE.md, and docs/1.0/ACCEPTANCE.md -- the current implementation contract.
- docs/adr/README.md -- the ADR index. Before touching a subsystem, read its ADR(s). ADRs are the authoritative record of every design decision; do not re-litigate a decided question, and do not silently contradict one. To change a decision, write a new ADR (or a marked amendment), never rewrite history.
- docs/SPEC.md -- the original design specification. Still the best deep explanation of the governance model, but ADRs supersede it where they differ.
- docs/DEV-LOOP.md -- read before any build/run/deploy work on a dev machine. It starts with a "when code changes, do this" table.
local/MACHINE-STATE.mdandlocal/NOTES.md-- if present (both gitignored): machine-local truth (which engine is running, install state, local gotchas) and sensitive/working memory (owner context, credential locations, session handoffs). Do not read them without explicit owner authorization. See local/README.md.- CONTRIBUTING.md -- test tiers, PR expectations, and the whole-repo Apache-2.0 OR MIT / DCO contribution terms.
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.
- 4d ago First seen · 231 lines · 3,293 tokens per session scan A e2f1f03fc011
ghostlight AGENTS.md is an instructions file published in the GitHub repository sylin-org/ghostlight (1 stars, last pushed yesterday), licensed Apache-2.0. It adds 3,293 tokens to every session, about $0.0165 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.
Other instructions, from other repositories
rustwright AGENTS.md
AGENTS.md instructions for Skyvern-AI/rustwright, covering repository instructions, public boundary, pull-request review, local verification and browser launches.
cloakbrowser-mcp AGENTS.md
Instructions for swimmwatch/cloakbrowser-mcp, covering agents.md, about the project, golden rules, project layout and daily commands.
mcp-stealth-chrome AGENTS.md
Instructions for RobithYusuf/mcp-stealth-chrome, covering mcp-stealth-chrome — agent guide, product and repository map, hard constraints, working defaults and architecture invariants.
cloakbrowser-mcp copilot-instructions.md
Instructions for swimmwatch/cloakbrowser-mcp: Follow AGENTS.md for project-wide coding, testing, documentation, changelog, and commit rules.
browser-mcp CLAUDE.md
Claude Code instructions for drisplabs/browser-mcp, covering claude.md, project overview, build and development commands, build and quality checks.
browser-mcp AGENTS.md
AGENTS.md instructions for drisplabs/browser-mcp, covering snapshot & interaction pipeline — audit results, critical (agent can't complete tasks), major (agent works around it poorly), m2. batch vs. fallback coordinate space mismatch and m3. wrapping hidden inputs dropped from snapshot.