Poteto's engineering rigor stack, adapted for Hermes Agent.
Fork of pstack by poteto (Meta/Netflix, React core team). Originally built as a Cursor plugin, adapted here for Hermes Agent with native skills, delegate_task orchestration, and playbook routing.
One orchestrator skill (poteto-mode) routes tasks to playbooks. Supporting skills cover exploration, design, review, delegation, cleanup, memory, teaching, verification, and workflow.
| Skill | Purpose |
|---|---|
pstack:poteto-mode |
Orchestrator and playbook routing. |
pstack:how |
Codebase exploration. |
pstack:architect |
Design types and module boundaries. |
pstack:interrogate |
Adversarial multi-reviewer review. |
pstack:swarm |
Parallel workers and aggregation. |
pstack:unslop |
Clean AI writing patterns from prose. |
pstack:deslop |
Clean AI code slop from diffs. |
pstack:setup-pstack |
Configure the delegate model and reasoning budget. |
pstack:automate-me, pstack:recall, pstack:reflect, pstack:show-me-your-work, pstack:why |
Personalization, history, reflection, evidence, and intent. |
pstack:benchmark-checklist, pstack:blast-radius, pstack:bro, pstack:create-verification-skill, pstack:figure-it-out, pstack:maintain-verification-skill, pstack:no-comments, pstack:tdd, pstack:teach, pstack:technical-writing, pstack:typescript-best-practices |
Focused engineering workflows. |
principle-* (20 skills) |
Engineering principles used by the orchestrator. |
Loaded via skill_view(name='pstack:poteto-mode', file_path='playbooks/<name>.md'):
investigation.md- read-only question, trace behaviorbug-fix.md- reproduce, root-cause, fix with evidencefeature.md- plan, design, delegate, verifyperf-issue.md- measure, profile, fix, measure againrefactoring.md- behavior-preserving restructuringautonomous-run.md- long task with falsifiable exit predicatefigure-it-out.md- large/cross-cutting work, design your own workflow
Read the guide in order the first time. After that, each page stands alone.
- Set up pstack - install, pick models, first task
- Route work through poteto-mode - give it a goal, watch it pick a playbook
- Understand the code - pstack:how before you edit anything
- Design the change - pstack:architect and pstack:interrogate before code locks in
- Build and clean - build playbooks, TDD, unslop
- Verify and ship - prove behavior, open a PR, merge
- Run work overnight - autonomous contracts, decision logs
- Principles - the full set in detail
- Make it yours - customize skills and playbooks
- Recipes and pitfalls - prompts to copy, mistakes to skip
Add the plugin's skills directory to skills.external_dirs so each skill registers as a slash command:
hermes config set skills.external_dirs '["plugins/pstack/skills"]'Then in any chat session:
| Command | What it does |
|---|---|
/poteto-mode |
Load the orchestrator (task routing, principles, delegation). |
/how |
Codebase exploration. "How does X work?" |
/architect |
Design types and boundaries before building. |
/interrogate |
Adversarial multi-reviewer code review. |
/swarm |
Fan out parallel workers. |
/unslop |
Strip AI writing patterns from prose. |
/setup-pstack |
Pick the delegate model and reasoning budget. |
Append an instruction after the command: /how how does auth work in this repo
Skills are also loadable programmatically via skill_view(name='pstack:<skill>').
Symlink or copy into your Hermes profile plugins directory:
ln -s /path/to/pstack ~/.hermes/profiles/<profile>/plugins/pstack
# Enable slash commands
hermes config set skills.external_dirs '["plugins/pstack/skills"]'Hermes loads it on next session start.
MIT