adr-advice
01Agent
Appends a piece of advice to the Advice section of an ADR. Invoke with: @adr-advice.
A Claude Code plugin marketplace for knowledge-management plugins — tooling that turns working knowledge into durable written artifacts
Agent
Appends a piece of advice to the Advice section of an ADR. Invoke with: @adr-advice.
Agent
Fills the Context section of a DRAFT ADR and updates the title to question form. Invoke with: @adr-context [contextnotes].
Agent
Completes an ADR by writing the Decision section, updating the title to a decision statement, and advancing Status to ACCEPTED. Invoke with: @adr-finalize.
Agent
Creates a new ADR file with empty sections and DRAFT metadata from the template. Invoke with: @adr-init.
Agent
Fills the Options Considered and Consequences sections of an ADR iteratively. Does not select an option. Invoke with: @adr-options.
Agent
Provisionally selects one option in an ADR, updates Consequences language, and advances Status to PROPOSED. Invoke with: @adr-propose.
Agent
Writes a fresh LLM-maintained knowledge base to disk from a ratified ontology spec — directory skeleton (raw/ concepts/ sources/ …), a GENERATED CLAUDE.md (page types, kind registry, relationship vocabulary, frontmatter schema, operations), README.md, empty index.md/log.md, .gitignore, the copied+injected operating…
Agent
Unifies ONE concept page that has accreted source-by-source. Lifts the cross-source synthesis up into ## What it is, demotes ## Across sources back to thin one-line-per-source provenance, and — for hub concepts — delegates re-explained detail down to the child pages that already own it. Source-voice only: never…