Curated list of Pyrite plugins. Plugins add custom entry types, MCP tools, CLI commands, validators, and presets.
These plugins ship in-tree in the extensions/ directory and are installable as separate packages.
Structured knowledge management for software teams -- ADRs, design docs, coding standards, component documentation, backlog tracking, and runbooks.
Install: pip install -e extensions/software-kb
Adds:
- Entry types:
adr,design_doc,standard,component,backlog_item,runbook - MCP tools:
sw_adrs,sw_component,sw_standards,sw_backlog,sw_create_adr,sw_create_backlog_item - CLI commands:
pyrite sw - Preset:
software - Workflows: ADR lifecycle, backlog workflow
- Relationships:
implements,supersedes,documents,depends_on,tracks(and inverses)
Use case: Engineering teams that want to keep architecture decisions, coding standards, component docs, and backlogs in version-controlled markdown alongside their code.
Personal knowledge management with atomic note-taking, CEQRC workflow, maturity tracking, and literature notes.
Install: pip install -e extensions/zettelkasten
Adds:
- Entry types:
zettel,literature_note - MCP tools:
zettel_inbox,zettel_graph - CLI commands:
pyrite zettel - Preset:
zettelkasten - Relationships:
elaborates,branches_from,synthesizes(and inverses)
Use case: Researchers, writers, and lifelong learners who practice Zettelkasten-style note-taking and want AI agents to help process their inbox of fleeting notes into permanent knowledge.
Wikipedia-inspired collaborative knowledge base with quality assessment, review workflows, protection levels, and talk pages.
Install: pip install -e extensions/encyclopedia
Adds:
- Entry types:
article,talk_page - MCP tools:
wiki_quality_stats,wiki_review_queue,wiki_stubs,wiki_submit_review,wiki_assess_quality,wiki_protect - CLI commands:
pyrite wiki - Preset:
encyclopedia - Workflows: Article review workflow
- DB tables: Encyclopedia review tables
Use case: Teams building shared reference documentation that needs editorial quality control -- quality ratings (stub through Featured Article), peer review queues, and page protection.
Everything2-inspired community knowledge base with user-authored writeups, voting, reputation tracking, and author-only editing enforcement.
Install: pip install -e extensions/social
Adds:
- Entry types:
writeup,user_profile - MCP tools:
social_top,social_newest,social_reputation,social_vote,social_post - CLI commands:
pyrite social - Preset:
social - Hooks:
before_save(author check),after_save(count updates),after_delete(reputation adjustment) - DB tables: Social vote and reputation tables
Use case: Communities that want a collaborative writing platform where members contribute essays, reviews, how-tos, and stories with upvote/downvote reputation mechanics.
Investigative journalism knowledge management covering actors, organizations, events, themes, mechanisms, scenes, victims, statistics, and timelines.
Install: pip install -e extensions/cascade
Adds:
- Entry types:
actor,cascade_org,cascade_event,timeline_event,theme,victim,statistic,mechanism,scene,solidarity_event - MCP tools:
cascade_actors,cascade_timeline,cascade_network,solidarity_timeline,solidarity_infrastructure_types,cascade_capture_lanes - KB types:
cascade-research,cascade-timeline,cascade-solidarity - Relationships:
member_of,investigated,funded_by,capture_mechanism,built_on,responded_to(and inverses)
Use case: Investigative journalists and researchers mapping networks of actors, organizations, events, and power structures across large-scale investigations.
Agent-oriented task management with a workflow state machine, parent-child decomposition, dependency tracking, evidence linking, and checkpoints.
Install: pip install -e extensions/task
Adds:
- Entry types:
task - MCP tools:
task_list,task_status,task_create,task_update,task_claim,task_decompose,task_checkpoint - CLI commands:
pyrite task - Preset:
task - Workflows: Task workflow (open, claimed, in_progress, blocked, review, done, failed)
- Hooks:
before_save(transition validation),after_save(parent rollup) - Relationships:
subtask_of,produces(and inverses)
Use case: AI agent orchestration -- agents claim tasks atomically, decompose work into subtasks, log checkpoints with confidence scores, and link evidence entries as they go.
Community plugins will be listed here. See Building Your Own to create and submit one.
Pyrite plugins are standard Python packages that expose entry types, MCP tools, CLI commands, validators, and presets through a simple class interface.
Resources:
- Plugin writing tutorial:
docs/tutorials/plugin-writing.md - Claude Code skill: Use the
extension-builderskill in Claude Code to scaffold a new plugin interactively - Reference implementations: Browse the extensions/ directory for working examples
Submitting a community plugin:
- Build and test your plugin against the latest Pyrite release
- Open a PR to markramm/pyrite adding your plugin to the Community Plugins section of this page
- Include a link to your plugin's repository, a one-line description, and what it adds