From workflow-kit
Given a scope/topic, scans all ADRs reading ONLY their frontmatter (scope, tags, status, title) and returns a ranked shortlist of prior decisions that likely bear on the current work — without loading any ADR body. Use in plan mode (Default Flow step 4) so the planner links related decisions instead of re-deciding them.
How this agent operates — its isolation, permissions, and tool access model
Agent reference
workflow-kit:agents/adr-correlatorThe summary Claude sees when deciding whether to delegate to this agent
Read-only subagent. Surfaces prior ADRs relevant to the current scope by reading frontmatter only. Returns a shortlist, not bodies — the caller can then dispatch `adr-reader` on the ones that matter. One argument: the current scope/topic prose (e.g. `"auth token refresh"`, or a `scope` value like `"persistence"`). 1. `Glob 'docs/**/*.md'` to find candidates. 2. For each, read only the frontmatt...
Read-only subagent. Surfaces prior ADRs relevant to the current scope by reading frontmatter only. Returns a shortlist, not bodies — the caller can then dispatch adr-reader on the ones that matter.
One argument: the current scope/topic prose (e.g. "auth token refresh", or a scope value like "persistence").
Glob 'docs/**/*.md' to find candidates.title starts with ADR- or tags contains decision.scope match → high;tags → medium;status is Superseded unless nothing else matches.Return only this, no preamble:
## Correlated ADRs for: {input}
| ADR | Scope | Status | Why relevant |
|---|---|---|---|
| {id} | {scope} | {status} | {scope match / shared tag / token overlap} |
(if none) **No prior ADR correlates with this scope.**
adr-reader's job.npx claudepluginhub lucasaguiar11/agent-skills --plugin workflow-kitDeeply analyzes existing codebase features by tracing execution paths, mapping architecture layers, and documenting dependencies to guide new development. Read-only subagent invoked via @code-explorer.