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 skills add gabrieldabbah/genesis --skill sourcesgit clone --depth 1 https://github.com/gabrieldabbah/genesisWrote 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/skills/gabrieldabbah/genesis/sources)<a href="https://agentmods.dev/skills/gabrieldabbah/genesis/sources"><img src="https://agentmods.dev/badge/skills/gabrieldabbah/genesis/sources.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 5.1 | $0.00137 | $0.01203 |
| Opus 5 | $0.00068 | $0.00602 |
| Sonnet 5 | $0.00027 | $0.00241 |
| Haiku 4.5 | $0.00014 | $0.00120 |
Grade A, and why
sources 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 7d 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 — 71 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Sources — trusted-first search & dependency vetting
Overview
A shown fact without a verified origin does not exist; an uncorroborated claim is a conjecture, not a fact. This skill applies that at the moment of searching and of choosing dependencies.
The registry is the project's docs/SOURCES.md; read it fresh each run — it is the source of truth for what to trust:
cat docs/SOURCES.md
sed -n '1,40p' CLAUDE.md # the stack and versions to pin to
Invocation is the authorization for the search/vetting itself. The only thing surfaced to the operator is a genuine fork (a license conflict, a risky-but-wanted dependency, a missing trusted source for a critical fact) — and the operator approves the actual install or commit, as for any change.
A — When asked to search / research / look something up
- Trusted-first. Identify the domain and go to its Tier-1 source in
SOURCES.md§3 before the open web (official language/framework docs, the standard, the package's own repo, this repo's own docs). - Version-pin. Match the project's installed major version (lockfile plus
CLAUDE.md) — docs for the wrong version are a silent trap. - Corroborate & cite. Any Tier-2/3 finding is confirmed against a Tier-1 source before you rely on it;
cite the origin inline. Ground every fact —
origin(fact) = None ⇒ drop it, never paraphrase a guess into a claim. - Conjecture-flag the rest. Anything from an unlisted or Tier-3 source is presented as a conjecture with its falsifier, or withheld — never asserted as fact.
- Compound the registry. If you found a genuinely trustworthy source not in §3, add it (tier + one
line on why) so the registry grows. For Claude/Anthropic model facts, defer to the
claude-apiskill.
B — When asked to add a dependency or tool
Run the SOURCES.md §4 vetting checklist — do not skip a box:
# real & canonical (not a typosquat): inspect the package's OWN registry page + linked repo
# e.g. npm view <pkg> repository.url / pip show <pkg> / cargo info <pkg> — then open the real repo
# advisory-clean, BOTH ENDS (the current version if replacing, and the target you'd add) — cmd from SOURCES.md §5:
# npm audit / pip-audit / cargo audit / govulncheck + query OSV / GitHub Advisory for exact name+version
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.
- 7d ago First seen · 71 lines · 137 tokens per session scan A 92176fee9ee4
sources is a skill published in the GitHub repository gabrieldabbah/genesis (4 stars, last pushed 1mo ago), licensed MIT. It adds 137 tokens to every session and 1,203 once invoked, about $0.0007 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 skills, from other repositories
fleet
Parallel campaign orchestrator. Runs multiple campaigns in coordinated waves within a single session. Spawns 2-3 agents per wave in isolated worktrees, collects discoveries, shares context between waves. Use when work decomposes into 3+ independent streams that can run simultaneously.
daemon
Continuous autonomous operation mode. Keeps campaigns running 24/7 by chaining Claude Code sessions via RemoteTrigger. Each session picks up from the campaign's continuation state, works until context runs low or the phase completes, then schedules the next session. Auto-stops on campaign completion or budget…
evolve
Research-driven multi-cycle improvement director. Forms causal hypotheses about why scores are low, validates them with scout agents before attacking, dispatches axis-parallel fleet attacks, extracts transferable patterns, and runs indefinitely within a budget envelope. Accumulates a persistent belief model and…
improve
Autonomous quality improvement loop. Scores a target against a rubric, selects the highest-leverage axis, attacks it, verifies, documents, and loops. No pre-planning between iterations — each loop re-scores from scratch.
setup
First-run experience for the harness. Three modes: Recommended (guided, 3 min), Full Tour (guided + skill walkthrough, 8 min), and Express (zero questions, 30 sec). Installs hooks first, detects stack, configures harness.json, runs a live demo on real code, and prints a reference card.
ascii-diagram
Generate perfectly aligned ASCII diagrams — architecture, flow, sequence, box-and-arrow. Uses a programmatic character-grid approach so alignment is guaranteed by math, not token prediction. Includes post-render verification.