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 DSB-117/brainblast --skill codex-skillgit clone --depth 1 https://github.com/DSB-117/brainblastWrote 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/dsb-117/brainblast/codex-skill)<a href="https://agentmods.dev/skills/dsb-117/brainblast/codex-skill"><img src="https://agentmods.dev/badge/skills/dsb-117/brainblast/codex-skill/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/dsb-117/brainblast/codex-skill"><img src="https://agentmods.dev/badge/skills/dsb-117/brainblast/codex-skill.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 2 findings, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high Prompt Injection · line 151 Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.Fix: Audit all comments and invisible characters. Remove any instructions that direct the agent to perform unauthorized actions. Use plain, reviewable content.
- medium Rogue Agent · line 33 Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.Fix: Remove any persistence mechanisms (cron jobs, startup scripts, state files). Skills should not maintain state across sessions without explicit user consent.
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 | $0.00043 | $0.04515 |
| Opus 5.5 | $0.00017 | $0.01806 |
| Sonnet 5.5 | $0.00009 | $0.00903 |
| Haiku 4.5 | $0.00004 | $0.00451 |
Grade A, and why
brainblast scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 426 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Brainblast
Research every external component in a requirements file before an agent starts coding. Produces .agent-research/runs/YYYYMMDD-HHMMSS/ with per-component notes and a final handoff report.
Incremental runs (caching). Research is cached per component, keyed by
name@version, in.agent-research/cache/. A re-run reuses cached components whose version is unchanged and re-researches only what changed. Components with no resolvable version are always re-researched. Pass--fresh(or setBRAINBLAST_FRESH=1) to ignore the cache. The cache is documentation only and safe to delete.
Preamble (run first)
# Browse setup — Codex gstack path
GSTACK_ROOT="$HOME/.codex/skills/gstack"
_REPO_ROOT=$(git rev-parse --show-toplevel 2>/dev/null)
[ -n "$_REPO_ROOT" ] && [ -d "$_REPO_ROOT/.agents/skills/gstack" ] && GSTACK_ROOT="$_REPO_ROOT/.agents/skills/gstack"
B="$GSTACK_ROOT/browse/dist/browse"
if [ -x "$B" ]; then
echo "BROWSE_READY: $B"
else
echo "BROWSE_MISSING"
fi
# Run directory + persistent component cache
_RUN_DIR="$(pwd)/.agent-research/runs/$(date +%Y%m%d-%H%M%S)"
mkdir -p "$_RUN_DIR/components"
_CACHE_DIR="$(pwd)/.agent-research/cache"
mkdir -p "$_CACHE_DIR"
_FRESH="${BRAINBLAST_FRESH:-0}" # set to 1 if the invocation included --fresh
_CI="${BRAINBLAST_CI:-0}" # set to 1 if the invocation included --ci
echo "RUN_DIR: $_RUN_DIR"
echo "CACHE_DIR: $_CACHE_DIR (fresh=$_FRESH)"
echo "DATE: $(date +%Y-%m-%d) (ci=$_CI)"
CI mode + gate. In --ci mode (_CI=1, or BRAINBLAST_CI=1) run non-interactively: never
ask a question or wait. On multiple requirement-file matches, pick the highest-precedence one
deterministically (requirements > prd > spec > brief > rfc > product > design-doc >
overview > scope > functional, then lexicographic) and say which; if none, stop BLOCKED and
write no report.json. Don't prompt for inventory confirmation. The pipeline's exit code comes from
the deterministic gate scripts/brainblast-gate.sh (Brainblast repo), which fails when a risk
at/above --fail-on (default critical) remains or the verdict is blocked; state PASS/FAIL
yourself after writing the report.
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 1mo ago First seen · 426 lines · 43 tokens per session scan A f663c41a7b73
brainblast is a skill published in the GitHub repository DSB-117/brainblast (93 stars, last pushed 3mo ago), licensed MIT. It adds 43 tokens to every session and 4,515 once invoked, about $0.0002 per session on Opus 5.5. 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 skills, from other repositories
e2e-alertmanager-test
Render end-to-end previews of what an alert notification will actually look like (plain-text email, HTML email, Slack attachment JSON, raw webhook JSON) by replaying a Prometheus unit-test file's expected alerts through a live Alertmanager. Effectively a "print preview" for alerts. Use to review notification…
alert-hysteresis
Query a live Prometheus's historical ALERTS series to statistically recommend better 'for:' hold-down durations, reducing spurious short-lived alerts and alert fatigue. Use interactively when an alert is firing too often/too briefly and needs its hysteresis tuned, or when auditing a rules file's for: values against…
autogen-promql-tests
Identify Prometheus alert/recording rules that have no promtool unit-test coverage, and scaffold true-positive/false-positive/hysteresis test-case skeletons for them. Use when bootstrapping unit tests for new or existing PromQL rules, or auditing test coverage of a rules file. Fully hermetic: reads only local…
label-check
Enforce that PromQL expressions (and, optionally, alert definitions) carry a required set of labels — e.g. job, namespace, cluster — to prevent label collisions in multi-tenant Prometheus/Thanos/Cortex/Mimir setups. Use when validating rule YAML for required labels or required alert labels, interactively or as a CI…
promql-fmt
Format and lint PromQL expressions embedded in Prometheus/Thanos/Cortex/Mimir rule YAML files — multiline layout, redundant aggregation-clause removal, aggregation-style consistency, and naming/instrumentation best-practice checks. Use whenever writing, editing, or reviewing alert/recording rule YAML, or as a CI…
stale-alerts-analyzer
Query a live Prometheus's historical ALERTS series to find alert rules that haven't fired in a long time — "dead code analysis" for alerting rules, surfacing candidates for deletion or threshold review. Use interactively during alert-quality/on-call hygiene reviews, not as a CI gate. Requires a live Prometheus with…