brainblast: Skill for Claude Code

.claude/skills/brainblast-fleet/SKILL.md

brainblast-fleet is a skill for Claude Code from DSB-117/brainblast. It costs 79 tokens per session (1,880 once invoked), scanned A, original, MIT.

An automated process for finding and verifying subtle problems in repositories that use a target software development kit (SDK), a package developers use to build against a service or platform.

In plain words
What is it for?
Use it to discover popular dependent repositories, send one scouting task to each, verify candidates, add proven findings to the corpus, and record them in a shared ledger.
Why use it?
It reduces manual searching and keeps unverified suggestions out of the shared collection. Proposed problems must be reproduced as a failing check and then fixed before promotion.

Skill for Claude Code

Written for Claude Code: allowed-tools in frontmatter. Also seen: mentions subagents.

This is DSB-117/brainblast's own configuration. It tells Claude Code how to work on brainblast itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything brainblast configures →

Reuse

Borrowing it

Nothing to install: this file belongs to DSB-117/brainblast. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/DSB-117/brainblast/main/.claude/skills/brainblast-fleet/SKILL.md
Clone the repo
git clone --depth 1 https://github.com/DSB-117/brainblast

Made for: Claude Code.

Wrote 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.

agentmods badge for brainblast-fleet

README.md
[![agentmods](https://agentmods.dev/badge/skills/dsb-117/brainblast/brainblast-fleet.svg)](https://agentmods.dev/skills/dsb-117/brainblast/brainblast-fleet)
Your own site
<a href="https://agentmods.dev/skills/dsb-117/brainblast/brainblast-fleet"><img src="https://agentmods.dev/badge/skills/dsb-117/brainblast/brainblast-fleet.svg" alt="Measured on agentmods" height="20"></a>
Per session 79 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,880 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. A grade says what 26 rules found in the file — not that it is safe.
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 5.1 $0.00079 $0.01880
Opus 5 $0.00039 $0.00940
Sonnet 5 $0.00016 $0.00376
Haiku 4.5 $0.00008 $0.00188

Measured 7d ago against content hash 71fd71a54d6c, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

brainblast-fleet scanned grade A with 1 finding 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.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

curl -s "${FLEET_REGISTRY_URL:-https://registry.brainblast.tech}/api/vti" | jq '.count, [.records[].id]'
.claude/skills/brainblast-fleet/SKILL.md · 165 lines

How it starts

The opening of the file, as written. The whole thing — 165 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Brainblast Fleet

The autonomous successor to hand-dropping candidates. You (the orchestrating agent) run this skill; it fans out a fleet of subagents — one per repository — to do the scouting, then the deterministic engine proves, promotes, submits, and logs. The model doing the reasoning is whatever agent is running Brainblast (this one) — no API key is requested.

The non-negotiable invariant: only traps that prove RED→GREEN through the existing checkers ever land. Subagents propose; proveFinding disposes.

Run from packages/core/. Five phases.

Phase 0 — Pick targets

If the user named SDKs/protocols, use them. Otherwise read the work-orders the last run left — cat ../../fleet/REPORT.md or datasets/COVERAGE.md — and target the uncovered classes and thin cells, freshness-first (recently-shipped or high-download SDKs, where models are most stale).

Phase 1 — Discover (deterministic)

For each target SDK:

npm run fleet:discover -- --sdk <pkg> --limit 10 --min-stars 200

Writes fleet/worklist.json (popular dependent repos, ranked by stars, already filtered against the shared ledger so you don't re-scout what another fleet did). Read it. If it's empty, the ledger says everything popular here is already done — pick another SDK.

Phase 2 — Scout (fan out the subagent fleet)

For each repo in the worklist, spawn a subagent (the Agent tool, general-purpose). Launch them in parallel batches (e.g. 3–5 at a time). Give each subagent EXACTLY this contract:

Scout <owner/repo> for a silent integration footgun in its use of <sdk>: code where the happy path compiles and runs but does the wrong thing (an insecure-default flag, a zeroed fee/amount, a skipped verification, a wrong constant). Shallow-clone read-only (git clone --depth 1) into a temp dir. For each real footgun you find that fits an existing brainblast checker (object-arg-property-forbidden-literal is the workhorse — an options-object property set to a forbidden string/number/boolean literal), write a candidate Finding to fleet/candidates/<id>.json following fleet/README.md. The vulnerable fixture must contain the forbidden value; the fixed fixture must set a safe literal (so it PASSES). Set class. Capture provenance from the real code — the registry requires it: record the exact git rev-parse HEAD of the shallow clone and, in the candidate's provenance, set sourceRef to github.com/<owner>/<repo>/blob/<that-sha>/<path> and evidence to the verbatim vulnerable line as it appears in the file. Without a real, fetchable sourceRef + evidence the finding can only enter the repo seed corpus, never the registry. Do not fabricate — if the repo has no provable footgun, write nothing and report "clean". Return: repo, the candidate ids you wrote (or "clean"), and a one-line note per finding.

Read the full file on GitHub · 165 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. 7d ago First seen · 165 lines · 79 tokens per session scan A 71fd71a54d6c

Subscribe to this mod's changes

brainblast-fleet is a skill published in the GitHub repository DSB-117/brainblast (100 stars, last pushed 1mo ago), licensed MIT. It adds 79 tokens to every session and 1,880 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.