#22977: fix(ui): resolve agent names from config in session dropdown

## Summary - **Problem:** The agent selection dropdown in the control dashboard showed mangled session keys like `agent-main-main (agent:main:main)` or delivery-context values (e.g., phone numbers) instead of the configured agent names (e.g., "Piper", "Clint"). - **Root cause:** `resolveSessionDisplayName` used `row.displayName` (populated from delivery context) as the preferred label and `parseSessionKey` returned the raw key as fallback for non-main agents — neither path looked up agent names from the loaded agents config. - **Fix:** Extract `agentId` from session keys, look up the agent's configured name from the agents list, and use it as the preferred display name (after user-editable labels but before row `displayName`). ## Change Type (select all) - [x] Bug fix - [ ] Feature - [ ] Refactor - [ ] Docs - [ ] Security hardening - [ ] Chore/infra ## Scope (select all touched areas) - [ ] Gateway / orchestration - [ ] Skills / tool execution - [ ] Auth / tokens - [ ] Memory / storage - [ ] Integrations - [ ] API / contracts - [x] UI / DX - [ ] CI/CD / infra ## Linked Issue/PR - Closes #22752 ## Changes **2 files changed** in `ui/src/ui/`: ### `app-render.helpers.ts` - Added `agentId?: string` field to `SessionKeyInfo` type - `parseSessionKey()` now extracts `agentId` from the second segment of `agent:X:…` keys - `resolveSessionDisplayName()` accepts an optional `agents: GatewayAgentRow[]` parameter; when an `agentId` resolves to a named agent in the config, that name is used as the display label (priority: user label → agent config name → row displayName → parsed fallback) - `resolveSessionOptions()` threads the agents list through to `resolveSessionDisplayName()` - `renderChatControls()` passes `state.agentsList?.agents` into `resolveSessionOptions()` ### `app-render.helpers.node.test.ts` - Updated existing `parseSessionKey` assertions to include the new `agentId` field - Added 2 new `parseSessionKey` tests for non-main agent keys - Added 9 new `resolveSessionDisplayName` tests covering: agent name resolution, priority over delivery-context displayName, user label priority, typed prefixes with agent names, fallback behavior when agent has no name / is not in list / agents list not provided ## User-visible / Behavior Changes With agents configured as: ```json "agents": { "list": [ { "id": "main", "name": "Piper" }, { "id": "clint", "name": "Clint" } ] } ``` | Session key | Before | After | |---|---|---| | `agent:main:main` | `agent-main-main (agent:main:main)` | **Piper** | | `agent:clint:main` | `+1(phone) (agent:clint:main)` | **Clint** | | `agent:main:subagent:abc` | `Subagent:` | **Subagent: Piper** | | `agent:main:cron:daily` | `Cron Job:` | **Cron: Piper** | User-editable session labels still take priority over agent names. ## Security Impact (required) - New permissions/capabilities? (`No`) - Secrets/tokens handling changed? (`No`) - New/changed network calls? (`No`) - Command/tool execution surface changed? (`No`) - Data access scope changed? (`No`) ## Test plan - [x] All 45 existing + new tests pass (`vitest run`) - [x] `parseSessionKey` returns `agentId` for all `agent:X:…` key patterns - [x] `resolveSessionDisplayName` prefers agent config name over delivery-context displayName - [x] User-editable labels still take priority over agent names - [x] Typed prefixes (Subagent:/Cron:) correctly applied to agent names - [x] Graceful fallback when agents list not available or agent not found - [ ] Manual verification in dashboard with multi-agent config 🤖 Generated with [Claude Code](https://claude.com/claude-code) <!-- greptile_comment --> <h3>Greptile Summary</h3> Fixed agent name display in session dropdown by extracting `agentId` from session keys and looking up configured agent names. Previously showed mangled keys or delivery-context values instead of friendly names like "Piper" or "Clint". The priority chain is now: user label → agent config name → row displayName → parsed fallback. <h3>Confidence Score: 5/5</h3> - Safe to merge - well-tested UI bug fix with comprehensive test coverage - Clean implementation with excellent test coverage (11 new tests). Logic is straightforward: extract agentId via regex, look up agent name, apply correct priority. No security concerns, purely UI-focused change. - No files require special attention _{Last reviewed commit: e674337} <!-- greptile_other_comments_section --> _{(2/5) Greptile learns from your feedback when you react with thumbs up/down!} <!-- /greptile_comment -->