feat(2157-1542): update CLAUDE.md agent table and delete research files (merge worktree-20260328-211544)

Author: JoeOakhartNava

.test-index

@@ -29,6 +29,7 @@ plugins/dso/scripts/gate-2b-blast-radius.sh: tests/scripts/test-gate-2b-blast-ra
 plugins/dso/scripts/detect-stack.sh:tests/scripts/test-detect-stack.sh
 plugins/dso/scripts/parse-template-registry.sh:tests/scripts/test-parse-template-registry.sh
 plugins/dso/config/template-registry.yaml:tests/scripts/test-template-validate-integration.sh
+config/agent-routing.conf: tests/scripts/test-agent-routing-conf.sh
 plugins/dso/scripts/validate-nava-platform-headless.sh:tests/scripts/test-validate-nava-platform-headless.sh
 plugins/dso/.claude-plugin/plugin.json:tests/plugin/test_fixture_minimal_plugin_consumer.py,tests/hooks/test-plugin-docs.sh,tests/hooks/test-hooks-json-paths.sh,tests/hooks/test-run-hook-relative-paths.sh,tests/scripts/test-plugin-retro-gather.sh,tests/scripts/test-dso-shim-plugin-root.sh,tests/scripts/test-plugin-scripts.sh,tests/scripts/test-collect-discoveries-plugin-root.sh,tests/scripts/test-plugin-scripts-no-relative-paths.sh,tests/scripts/test-check-plugin-test-needed.sh,tests/scripts/test-plugin-dir-structure.sh,tests/scripts/test-plugin-reference-catalog.sh,tests/scripts/test-no-unguarded-plugin-refs.sh
 plugins/dso/hooks/cascade-circuit-breaker.sh:tests/hooks/test-cascade-breaker.sh
@@ -119,6 +120,8 @@ plugins/dso/skills/fix-bug/prompts/escalated-investigation-agent-3.md:tests/skil
 plugins/dso/skills/fix-bug/prompts/escalated-investigation-agent-4.md:tests/skills/test_escalated_investigation_agent_4_prompt.py
 plugins/dso/skills/fix-bug/prompts/intermediate-investigation.md:tests/skills/test_intermediate_investigation_prompt.py
 plugins/dso/skills/fix-bug/SKILL.md:tests/skills/test_fix_bug_skill.py
+plugins/dso/skills/fix-bug/SKILL.md:tests/skills/test_fix_bug_gate_1a.py [test_gate_1a_position]
+plugins/dso/skills/fix-bug/SKILL.md:tests/skills/test_fix_bug_gate_1b.py [test_gate_1b_pre_investigation]
 plugins/dso/skills/fix-bug/prompts/intermediate-investigation-fallback.md:tests/skills/test_fix_bug_skill.py
 plugins/dso/skills/fix-bug/prompts/basic-investigation.md:tests/skills/test_basic_investigation_prompt.py
 plugins/dso/skills/fix-bug/prompts/intermediate-investigation.md:tests/skills/test_intermediate_investigation_prompt.py
@@ -210,6 +213,7 @@ plugins/dso/skills/preplanning/SKILL.md: tests/skills/test-scrutiny-pending-gate
 plugins/dso/skills/implementation-plan/SKILL.md: tests/skills/test-scrutiny-pending-gate.sh
 CLAUDE.md: tests/skills/test-shared-subworkflow-docs.sh
 plugins/dso/agents/doc-writer.md: tests/scripts/test-doc-writer-agent.sh
+plugins/dso/agents/bot-psychologist.md: tests/scripts/test-bot-psychologist-agent.sh
 plugins/dso/skills/update-docs/SKILL.md: tests/hooks/test-sub-agent-guard.sh
 # overlay dispatch logic (task 9937-4de6 / story w22-25ui) — no single source file yet (implemented by task 67e2-2912)
 plugins/dso/scripts/overlay-dispatch.sh: tests/hooks/test-overlay-dispatch.sh

CLAUDE.md

@@ -66,6 +66,7 @@ Priority: 0-4 (0=critical, 4=backlog). Never use "high"/"medium"/"low".
 |-------|-------|---------------|
 | `dso:complexity-evaluator` | haiku | `/dso:sprint`, `/dso:brainstorm`; read inline by `/dso:fix-bug` |
 | `dso:conflict-analyzer` | sonnet | `/dso:resolve-conflicts` |
+| `dso:bot-psychologist` | sonnet | `/dso:fix-bug` llm-behavioral path (dispatched or read inline when sub-agent) |
 | `dso:doc-writer` | sonnet | `/dso:sprint` (doc stories), `/dso:update-docs` |
 | `dso:intent-search` | sonnet | `/dso:fix-bug` Step 1.5 (Gate 1a — pre-investigation intent search) |
 | `dso:feasibility-reviewer` | sonnet | `/dso:brainstorm` (conditional, on integration signals) |

config/agent-routing.conf

@@ -36,8 +36,13 @@
 #   error-debugging:error-detective               | error-debugging@claude-code-workflows
 #   code-simplifier:code-simplifier               | code-simplifier@claude-plugins-official
 #   backend-api-security:backend-security-coder   | backend-api-security@claude-code-workflows
+#   dso:bot-psychologist                          | (local plugin agent, resolved via plugins/dso/agents/)
 #   general-purpose                               | (built-in, no plugin required)
 #
+# Note: dso: agents are local plugin agents resolved via plugins/dso/agents/<name>.md.
+# They are resolved as always-available by discover-agents.sh without requiring an
+# enabledPlugins entry (short-circuit path at agent resolution loop).
+#
 
 # ── Routing Table ────────────────────────────────────────────────────────────
 
@@ -48,3 +53,4 @@ mechanical_fix=debugging-toolkit:debugger|code-simplifier:code-simplifier|genera
 complex_debug=error-debugging:error-detective|debugging-toolkit:debugger|general-purpose
 code_simplify=code-simplifier:code-simplifier|general-purpose
 security_audit=backend-api-security:backend-security-coder|general-purpose
+llm_behavioral=dso:bot-psychologist|general-purpose