Menú
7-persona engineering safety council with veto authority. Trigger phrase: "technical working group".
Instalar con Codex o Claude Copia este prompt, pégalo en Codex, Claude u otro asistente, y deja que revise la página de la skill y la instale por ti.
Basado en la clasificación ocupacional SOC
Every 2h routes unclaimed kanban tasks to agent skills for execution
Morning kanban review — parse all agent carryovers after cron reports, surface open questions as Hermes kanban tasks, update agent status, deliver Discord summary. Creates tasks in hermes kanban (not sheet.md). Proactive: each morning agent cron triggers its own carryover parse.
ArXiv research curator agent — bootstraps from agent-sheets/arxiv.md directives. Discovers, downloads, and reports on ML/AI papers from arXiv.
Ingest pipeline operator agent — bootstraps from agent-sheets/ingest.md directives. Processes raw files from raw/ inbox into structured wiki knowledge.
Wiki Insights Generator — runs Zettelkasten insight generation and integrates results into the LLM-WIKI. Runs daily at 06:00 AM.
Kanban board creation and workflow decomposition persona. Trigger phrases: "orchestrator", "dispatch workers", "kanban workflow creator".
Runtime skill: ~/.hermes/skills/autonomous-ai-agents/technical-working-group/SKILL.md
The Technical Working Group is an engineering safety council. Every proposal that passes the Research Council comes here for feasibility review and veto if needed.
"technical working group" — activates the engineering council"twg" or "engineering council" — shorthand| Persona | Specialty | Veto Trigger |
|---|---|---|
| Requirements | Scope and definition | Ambiguous requirements |
| Dev | Implementation | Unsafe code paths |
| QA | Testing and verification | Insufficient test coverage |
| Security | Attack surface | Undefined or unbounded input |
| Data | Data integrity | Schema drift, data loss risk |
| Performance | Bottlenecks and scale | Exponential complexity |
| Observability | Debugging and tracing | No failure visibility |
| Refuser | Veto authority | Named + plausible + non-reversible |
The Refuser holds the deploy token and can veto any work item that meets all three conditions:
If a veto fires: work item is blocked, orchestrator is notified, synthesis is updated.
The runtime skill lives in ~/.hermes/skills/autonomous-ai-agents/technical-working-group/SKILL.md. No installation needed.
hermes profile chat --profile technical-working-group
Or trigger via orchestrator dispatch.
The TWG reviews work items from the kanban board. Each persona evaluates:
## TWG Review
- Requirements: ✅ Approved
- Dev: ⚠️ Concerns about error handling
- Security: ❌ VETO — unbounded input on /api/process
- Refuser: VETO — named workers, plausible overwork, non-reversible burnout
Approved items return to orchestrator for worker dispatch.
IF named AND plausible AND non-reversible:
VETO = True
deploy_token.revoke()
notify(orchestrator, veto_reason)
The Refuser is the last line of defense. If in doubt, the Refuser vetoes.
~/.hermes/profiles/technical-working-group/SOUL.md — identity and directive~/.hermes/skills/autonomous-ai-agents/technical-working-group/SKILL.md — runtime execution