Skills

Browse public skills registered by Harnez. Search matches names, purpose, source, requirements, targets, and invocation.

31 skills

first-party · AGPL-3.0-or-later

docup

Review and update bounded repository documentation by category (testing, architecture)

bundled · MIT

domain-modeling

Build and sharpen a project's domain model. Use when discussing codebase terminology, writing or editing a CONTEXT.md, or recording or editing an ADR.

first-party · AGPL-3.0-or-later

evergreen

Create/update evergreen docs and issue files for the current session

bundled · MIT

grill-with-docs

A relentless interview to sharpen a plan or design, which also creates docs (ADR's and glossary) as we go.

bundled · MIT

grilling

Grill the user relentlessly about a plan, decision, or idea. Use when the user wants to stress-test their thinking, or uses any 'grill' trigger phrases.

first-party · AGPL-3.0-or-later

harnez-advisor

Reuse compatible advisor sessions across supported agent tools using prose requests

first-party · AGPL-3.0-or-later

harnez-handoff

Write ready-to-paste handoff prompts for a local or cloud agent, optionally exploring for valuable work first

first-party · AGPL-3.0-or-later

harnez-init

Re-run harnez init in an initialized repository, assess the diff for regressions, fix local fallout, and commit changes

first-party · AGPL-3.0-or-later

harnez-selftest

Test and verify native agent background task execution flow and self-test command progression

first-party · AGPL-3.0-or-later

harnez-status

Briefly self-check Harnez CLI, model tiers, sprint workflow, and documentation navigation

first-party · AGPL-3.0-or-later

issue

File a new repository issue or ticket when asked; do not trigger merely because work references an existing issue

first-party · AGPL-3.0-or-later

lean-sprint

Execute a lean fresh-handoff sprint (Goal Handoff -> Autonomous Dev -> Inline Review -> Hygiene)

first-party · AGPL-3.0-or-later

lean-sprinter

Delegate a lean fresh-handoff sprint to a user-selected-model subagent that reads and executes the lean-sprint skill

first-party · AGPL-3.0-or-later

lmcoder

Local LLM server lifecycle management, fast interactive prompt streaming with session continuity, and sandboxed coding agent execution

first-party · AGPL-3.0-or-later

mode

Switch ConciseMode terseness level (lite, std, ultra, off) and sync AGENTS.md

first-party · AGPL-3.0-or-later

Continue the most recent workflow on its next actionable item

first-party · AGPL-3.0-or-later

peer-assistant

Coordinate peer requests and ticket work across agent sessions

first-party · AGPL-3.0-or-later

Multi-project release & publish pipeline: detects project shape (CLI, artifact app, website, hybrid), executes pre-flight checks, verifies media, requires explicit confirmation before external push, syncs website, and asserts live endpoint liveness

first-party · AGPL-3.0-or-later

release

Lean CLI release: commit pending code, push, run harnez release, fix issues or ask on login

first-party · AGPL-3.0-or-later

respect

Move hardcoded values that belong in a repository's embedded YAML spec into the spec

first-party · AGPL-3.0-or-later

reverse-sprint

Execute a reverse sprint (Low-Cost Dev Lead -> Frequent Auto-Compact -> Milestone Reviews -> Bounded Advisor Escalation)

first-party · AGPL-3.0-or-later

reverse-sprinter

Delegate a reverse sprint to a low-cost developer agent that reads and executes the reverse-sprint skill

first-party · AGPL-3.0-or-later

review

Review current changes, verify issue completion if specified, check code quality, apply small fixes directly, and file follow-up issues for larger findings

first-party · AGPL-3.0-or-later

roadmap

Product-manager-style roadmap creation/update from the open issue backlog (check issues -> check existing roadmap -> synthesize/reconcile by product value axis)

first-party · AGPL-3.0-or-later

sprint

Orchestrate a 5-phase agentic sprint loop (Sequential Advisor -> Reusable Dev -> Review -> Hygiene -> Retro)

first-party · AGPL-3.0-or-later

sprinter

Delegate a full 5-phase sprint to a user-selected-model subagent that reads and executes the sprint skill

first-party · AGPL-3.0-or-later

story

Write a candid engineering retrospective and case study in docs/studies/

bundled · MIT

to-spec

Turn the current conversation into a spec and publish it to the project issue tracker: no interview, just synthesis of what you've already discussed.

first-party · AGPL-3.0-or-later

tool-feedback-protocol

Use after an internal tool call (Read, Edit, Bash, etc.) in a harnez-tracked project (issues/ present) failed or didn't produce the expected outcome — record it via harnez rate before moving to the next step. Also use if a harnez tip nudges you after a long quiet stretch and nothing has failed — confirm with harnez rate --ok. Skip both for routine successful calls.

first-party · AGPL-3.0-or-later

website

Build or update a project website — reads docs/Website.md rules (hosting, branding, static/no-CDN, subpage links) before generating any content

first-party · AGPL-3.0-or-later

wrap

Close the current work session, preserve its state, and prepare a concise handoff