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 martineserios/thebrana --skill brainstormgit clone --depth 1 https://github.com/martineserios/thebranaWrote 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/martineserios/thebrana/brainstorm)<a href="https://agentmods.dev/skills/martineserios/thebrana/brainstorm"><img src="https://agentmods.dev/badge/skills/martineserios/thebrana/brainstorm.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 | $0.00032 | $0.06306 |
| Opus 5 | $0.00016 | $0.03153 |
| Sonnet 5 | $0.00006 | $0.01261 |
| Haiku 4.5 | $0.00003 | $0.00631 |
Grade A, and why
brainstorm 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 — 600 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Brainstorm
Interactive idea maturation. You have a rough idea — this skill helps you explore it, research what exists, challenge assumptions, and shape it into something concrete.
Usage
/brana:brainstorm [idea]
- With idea: start exploring immediately
- Without idea: ask what's on your mind
Principles
- The user leads. They own the idea. You research, challenge, and structure — never hijack.
- Interactive always. Every phase uses AskUserQuestion for direction. Never monologue.
- Research proactively. When a tool, framework, or concept is mentioned, auto-fetch context.
- Diverge then converge. First widen (explore angles), then narrow (pick direction).
- Concrete over abstract. Push toward specifics — who, what, how, when.
Step Registry
On entry, create a CC Task step registry. Follow the guided-execution protocol.
Register these steps: LOAD, SEED, EXPAND, DISCUSS, SHAPE, OUTPUT, EXTRACT, EVALUATE, PERSIST.
ToolSearch("select:mcp__ruflo__memory_search,mcp__brana__agy_delegate")
Procedure
Step 0 — LOAD
Pull relevant knowledge into context before the brainstorm begins. Budget: 30K tokens max.
- Build query from available context:
"{project} {task.subject} {task.tags joined} {user_input}" - Primary — ruflo MCP (run both in parallel —
namespace: "all"only returns session records;specsnamespace is unindexed):
Merge results, rank by similarity. Focus on: dimension docs, idea docs (mcp__ruflo__memory_search(query: "{query}", namespace: "knowledge", limit: 4, threshold: 0.4) mcp__ruflo__memory_search(query: "{query}", namespace: "pattern", limit: 3, threshold: 0.4)docs/ideas/), and recent research findings. 2b. Graph edge traversal — seebuild.mdLOAD step 2b. Followdepends_on/informsedges from knowledge results. Max 3 graph-derived docs. Best-effort, never blocks. - Fallback — tag-based grep (if MCP unavailable):
Read the top 3 matching files (first 80 lines each).grep -rl "{keywords}" ~/enter_thebrana/brana-knowledge/dimensions/ --include="*.md" | head -5 grep -rl "{keywords}" docs/reflections/ docs/ideas/ --include="*.md" | head -5 - Skill match handling — if any result has
namespace: "skills"and score >= 0.5, mention inline: "Matching skill: /brana:{name} ({score})." Informational only — don't auto-invoke or block. 4a. JIT skill acquisition — if no skills match and topic involves a specific technology, offer marketplace search viaSkill(skill="brana:acquire-skills", args="{tech}"). Read installed procedure into context immediately. Seebuild.mdLOAD step 4a for full logic and guard rails. - Summarize loaded knowledge as a brief context preamble (2-5 bullets). Do not show raw results — synthesize what's relevant to the seed idea.
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 · 600 lines · 32 tokens per session scan A 7442333d0656
brainstorm is a skill published in the GitHub repository martineserios/thebrana (3 stars, last pushed today), licensed MIT. It adds 32 tokens to every session and 6,306 once invoked, about $0.0002 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 2026-09-03.
Other skills, from other repositories
bump-catalog
Promote frequently used mojiemoji variants from the local usage cache into prestamp-catalog.yml and open a deterministic catalog update pull request.
mojiemoji-config
Manage persistent mojiemoji prestamp settings such as intensity in /.config/mojiemoji/config.json.
mojiemoji-github
GitHub issue / PR / review / release bodies for Codex should run through the shared mojiemoji core before posting Japanese Markdown.
eu-ai-act-check
EU AI Act compliance — migrated to pitimon/claude-governance v3.1.0. This stub redirects to the canonical implementation. Maps to H1 (Be Proactive — prevent regulatory crisis) + H8 (Voice/Conscience).
workflow
Guided walkthrough of the 7-step workflow. Prompts at each step to invoke or skip. Use when starting a new feature or when unsure which step comes next.
log
A phase-log command that records completed work in Markdown files, a human-readable text format, and updates an index of phases.