Skip to main content

Plugins reference

This page lists the 17 plugins of the Navin library. Activating a plugin in a project activates its commands, agents and skills.

NameVersionCategoryCommandsAgentsSkillsDescription
agent-sdk-dev1.0.0Development120Scaffold and verify apps built on the Navin agent: Python on the Navin SDK, TypeScript on an OpenAI-compatible client pointed at navin serve, with current versions checked, a type check and a PASS / PASS WITH WARNINGS / FAIL review.
code-review1.0.0Productivity112Pull request review by parallel agents with validation of every finding, so only high-signal issues are reported or posted inline.
commit-commands1.0.0Productivity302Git workflow commands: commit, commit + push + pull request, and cleanup of local branches deleted on the remote.
document-skills1.0.0Documents008Office document suite: read, create, edit, validate and convert Word (.docx), Excel (.xlsx), PowerPoint (.pptx) and PDF files, with tracked changes, comments, formula recalculation, slide thumbnails and PDF forms.
example-skills1.0.0Design0012Creative and builder pack: generative art, Navin brand guidelines, canvas and poster design, ready-made themes, Slack GIFs, document co-authoring, internal communications, MCP server building, web app testing, single-file web artifacts and skill creation with an evaluation loop.
explanatory-output-style1.0.0Learning101The agent explains as it codes: short Insight blocks about implementation choices and codebase patterns, injected into the system prompt every turn.
feature-dev1.0.0Development133Guided feature development: codebase exploration, clarifying questions, competing architectures, implementation and parallel quality review.
frontend-design1.0.0Development003Distinctive, production-grade interfaces that avoid the generic generated look: subject-driven direction, typography, palette, restrained motion and interface copy, built inside the locked design system.
hookify1.0.0Productivity412Warn about or block unwanted agent behaviour with simple Markdown rules (.navin/hookify.<name>.local.md), created from an instruction or from the conversation, checked natively on every action.
infra-ops1.0.0Devops0027Cloud, DevOps, SRE and platform engineering bench driven by the infra engine: discover, graph, plan, run through the broker, verify and recover. Entry point: infra-orchestrator.
issue-triage1.0.0Productivity202Issue maintenance: find likely duplicates and comment them, triage an issue by applying only labels that exist.
learning-output-style1.0.0Learning101Learn by doing: the agent hands meaningful decisions to you with TODO(human) markers, does the boilerplate itself, and adds Insight blocks.
model-migration1.0.0Development101Migrate an app's code, config and prompts to a newer model generation for any provider: inventory model ids, update ids and parameters (effort, thinking budget, limits, retired fields), tune prompts for more literal models, compare before and after.
plugin-dev1.0.0Development137Toolkit to build Navin plugins: skills, agents, slash commands, hooks, MCP servers and settings, with guided creation and validation.
pr-review-toolkit1.0.0Productivity162Specialised review agents for comments, tests, error handling, type design, code quality and simplification, orchestrated by /review-pr.
ralph-loop1.0.0Productivity300Self-improving iteration loop: the same task comes back at every end of turn, so the agent keeps refining its own work until an honest completion promise or an iteration cap.
security-guidance1.0.0Security111Security reminders when an edit adds a dangerous pattern (eval, pickle, shell=True, raw innerHTML, disabled TLS...), an opt-in AI review of the diff at end of turn and after each commit, and /security-review on demand.