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 commands/quantulabs/hivemind/hivegit clone --depth 1 https://github.com/QuantuLabs/HivemindWhat 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 | $0.00000 | $0.01446 |
| Opus 5 | $0.00000 | $0.00723 |
| Sonnet 5 | $0.00000 | $0.00289 |
| Haiku 4.5 | $0.00000 | $0.00145 |
Grade A, and why
hive 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 yesterday.
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 — 209 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/hive - Hivemind Deliberation
Query multiple AI models and orchestrate consensus through deliberation.
You (Claude Code) are the orchestrator. GPT and Gemini provide their perspectives, and you analyze, investigate, and synthesize the final consensus.
Arguments
$ARGUMENTS- The question to ask the Hivemind
Phase 1: Context Building (if needed)
For questions about the codebase:
CRITICAL: Pass RAW, UNDIGESTED code - but be SMART about what you include.
- Use Glob to find relevant files by pattern
- Use Read to get file contents
- Be selective - include the relevant portions, not necessarily entire files
- DO NOT use Explore agents - they return analyzed summaries, not raw code
- DO NOT pre-analyze - keep your interpretation for AFTER hivemind results
Smart context selection:
- For audits: Include the actual implementation code, not your analysis
- For bugs: Include the buggy function + related code, line numbers
- For architecture: Include key files/interfaces, skip boilerplate
- Large files: Use Read with offset/limit to get relevant sections
Example workflow:
# Find files
Glob("src/auth/**/*.ts")
# Read targeted sections (smart selection)
Read("/path/to/auth.ts", offset=50, limit=100) # Just the relevant function
Read("/path/to/types.ts") # Full file if small and relevant
# Pass raw code to hivemind (next phase)
Key: Raw ≠ Everything. Raw = actual code, not your summary of it.
For general questions, skip to Phase 2.
Phase 2: Initial Query
Step 1: Formulate YOUR answer first
Before calling the hivemind tool, think through the question and formulate your own complete answer. Be thorough - this is your contribution to the deliberation.
Step 2: Query other models
hivemind({
question: "The question here",
context: `
## Source Code (RAW - not summaries)
### /path/to/auth.ts (lines 50-150)
\`\`\`typescript
[PASTE THE RELEVANT CODE SECTION - actual code, not your description]
\`\`\`
### /path/to/types.ts
\`\`\`typescript
[PASTE ACTUAL CODE - be selective if file is large]
\`\`\`
`
})
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.
- yesterday First seen · 209 lines · 0 tokens per session scan A 4b3f82d8217f
hive is a command published in the GitHub repository QuantuLabs/Hivemind (1 stars, last pushed 7mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,446 tokens. 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 commands, from other repositories
checklist
Generate a custom checklist for the current feature based on user requirements.
clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.
specify
Create or update the feature specification from a natural language feature description.
analyze
Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.
converge
Assess the current codebase against the feature's spec, plan, and tasks, then append any remaining unbuilt work as new tasks to tasks.md so implement can complete it.
implement
Execute the implementation plan by processing and executing all tasks defined in tasks.md.