docs: fix S1, S2, S3 from review — context passing, agent types, frontmatter

S1: code-reviewer passes Phase 1 findings to specialists in task brief
    so they skip re-querying dev_refs/dev_map
S2: research-planner clarifies general-purpose is a Claude Code built-in
    subagent type (default when no subagent_type specified)
S3: Part 2.4 specifies exact tools: frontmatter changes for each agent

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Author: prosdev

.claude/agents/code-reviewer.md

@@ -39,19 +39,27 @@ Based on what you learned, write **specific focused tasks** for each specialist.
 Do NOT send them the same generic "review the diff" prompt. Tell each one exactly
 what to focus on.
 
+**Include your Phase 1 findings in each task brief.** The specialists have MCP tools
+too, but they should NOT re-query what you already gathered. Pass them the callers,
+hot path rank, and pattern context you found — so they focus on judgment, not data
+gathering.
+
 Example — bad (generic):
 > "security-reviewer: review the diff for security issues"
 
-Example — good (focused):
+Example — good (focused, includes Phase 1 context):
 > "security-reviewer: This PR adds a new `resolveTarget` function that runs
 > `execSync('git diff ...')` with user-provided input at refs.ts:67. Check for
 > command injection. Also review the new `graphPath` config that's passed from
-> user config to fs.readFile at review-analysis.ts:42."
+> user config to fs.readFile at review-analysis.ts:42.
+>
+> Context from my analysis: this file has 12 downstream consumers (high impact).
+> The execSync call is the only shell invocation in the diff."
 
 Write focused tasks for:
-- **security-reviewer** — point it at specific user input paths, shell commands, file access
-- **logic-reviewer** — point it at specific error handling, race conditions, edge cases you spotted
-- **quality-reviewer** — point it at specific test gaps, naming inconsistencies, convention deviations
+- **security-reviewer** — point it at specific user input paths, shell commands, file access. Include which files are high-impact.
+- **logic-reviewer** — point it at specific error handling, race conditions, edge cases. Include the callers/callees you found.
+- **quality-reviewer** — point it at specific test gaps, naming inconsistencies. Include the pattern comparison results.
 
 ### Phase 3: Delegate in parallel
 

.claude/agents/research-planner.md

@@ -66,8 +66,10 @@ Plan 2-4 research tasks. Each should:
 
 ### Phase 3: Delegate research in parallel
 
-Launch sub-agents via the Agent tool, one per research task. Use the
-`general-purpose` agent type. Give each a precise brief:
+Launch sub-agents via the Agent tool, one per research task. These use
+Claude Code's built-in `general-purpose` subagent type (not a custom agent
+definition — it's the default when no `subagent_type` is specified).
+Give each a precise brief:
 
 ```
 Agent 1: "Search GitHub for how [specific project] implements [specific thing].

.claude/da-plans/mcp/phase-2-composite-tools/2.4-docs-and-agents.md

@@ -39,19 +39,49 @@ Update CLI commands section to include `dev review` and `dev research`.
 
 Update Agent → MCP Tool Matrix to show which agents benefit from composite tools.
 
-### Agent updates
+### Agent updates — specific frontmatter changes
 
-**code-reviewer.md** — Use `dev_review` instead of manually orchestrating 3 reviewers.
-The `dev_review` MCP tool gives each sub-agent the context it needs in one call.
+**code-reviewer.md** — Add `mcp__dev-agent__dev_review` to `tools:` frontmatter.
+Use `dev_review` in Phase 1 to gather impact/patterns/structure in one call
+instead of calling `dev_refs` + `dev_map` + `dev_patterns` separately.
 
-**research-planner.md** — Use `dev_research` as the first step to map the territory.
-Replaces the manual `dev_map` + `dev_search` + `dev_patterns` workflow.
+```yaml
+# Current tools:
+tools: Read, Grep, Glob, Bash, mcp__dev-agent__dev_search, mcp__dev-agent__dev_refs, mcp__dev-agent__dev_map, mcp__dev-agent__dev_patterns
+# Add:
+tools: ..., mcp__dev-agent__dev_review
+```
+
+**research-planner.md** — Add `mcp__dev-agent__dev_research` to `tools:` frontmatter.
+Use `dev_research` in Phase 1 to map the territory in one call.
+
+```yaml
+# Current tools:
+tools: Read, Grep, Glob, Bash, mcp__dev-agent__dev_search, mcp__dev-agent__dev_refs, mcp__dev-agent__dev_map, mcp__dev-agent__dev_patterns
+# Add:
+tools: ..., mcp__dev-agent__dev_research
+```
 
-**bug-investigator.md** — Use `dev_research "error in <component>"` to quickly gather
-all relevant context for a bug investigation.
+**bug-investigator.md** — Add `mcp__dev-agent__dev_research` to `tools:` frontmatter.
+Use `dev_research "error in <component>"` to quickly gather all relevant context.
 
-**plan-reviewer.md** — Use `dev_review` to verify that plan claims match actual
-codebase state.
+```yaml
+# Current tools:
+tools: Read, Edit, Write, Glob, Grep, Bash, mcp__dev-agent__dev_search, mcp__dev-agent__dev_refs, mcp__dev-agent__dev_map
+# Add:
+tools: ..., mcp__dev-agent__dev_research
+```
+
+**plan-reviewer.md** — Add `mcp__dev-agent__dev_review` to `tools:` frontmatter.
+Use `dev_review` to verify that plan claims about file impact and conventions
+match actual codebase state.
+
+```yaml
+# Current tools:
+tools: Read, Grep, Glob, Bash, mcp__dev-agent__dev_refs, mcp__dev-agent__dev_map, mcp__dev-agent__dev_patterns
+# Add:
+tools: ..., mcp__dev-agent__dev_review
+```
 
 ### `website/content/updates/index.mdx`