add-eval
Add or update an LLM eval test for a specific tool or collection. Creates a new eval scenario or updates an existing one to cover new tools. Use when adding eval coverage for new or modified tools.
What this skill does
# Add Eval
Add or update an LLM eval test for a specific tool or collection. This skill creates a new eval scenario or updates an existing one to cover new tools.
Use this skill when:
- You added a new tool with `/add-tool` and want eval coverage
- You modified a tool and need to update the eval scenario
- You want to add a new workflow scenario to an existing collection
- `/build-evals` skipped the collection because eval files already exist
## Prerequisites
Before running, ensure:
1. The collection exists (`src/umbraco-api/tools/{collection}/index.ts`)
2. The project builds: `npm run build`
3. An LLM API key is available (eval tests use Claude Agent SDK — works with a Claude Code subscription or `ANTHROPIC_API_KEY`)
## Arguments
- **Collection name** (required): the collection to add eval coverage for (e.g. `/add-eval form`)
- **Tool or scenario hint** (optional): what to test (e.g. `/add-eval form "copy form workflow"`)
If no hint is provided, compare the tools in the collection against existing eval test coverage and suggest what's missing.
## Agents
| Agent | When to use |
|-------|-------------|
| `eval-test-creator` | Creating or updating eval test files (Step 3 or 4) |
## Critical Rules
**BUILD BEFORE RUNNING.** Eval tests run against `dist/index.js`. Always `npm run build` first.
**RUN COMMANDS SEPARATELY.** Always run build and test as separate Bash calls. Never chain with `&&`.
**ITERATE ON PROMPTS.** Eval tests are probabilistic. If a test fails, the fix is usually in the prompt.
**VERBOSE DURING DEVELOPMENT.** Always set `verbose: true` when creating or debugging.
## Workflow
### Step 0: Check Eval Infrastructure
Check if the eval setup exists:
- `tests/evals/helpers/e2e-setup.ts`
- `tests/evals/jest.config.ts`
If the setup doesn't exist, tell the user to run `/build-evals {collection}` first to create the infrastructure. This skill does not create eval setup files.
### Step 1: Understand Current Coverage
Read the collection's tools and existing eval tests:
- `src/umbraco-api/tools/{collection}/index.ts` — all tools in the collection
- `tests/evals/{collection}-*.test.ts` — existing eval scenarios
- Each eval test file — to understand which tools are covered and in what scenarios
Build an inventory:
- Tools with eval coverage (appear in a scenario's `tools` and `requiredTools` arrays)
- Tools without eval coverage
- Existing workflow scenarios
### Step 2: Decide: Update or Create
Based on the inventory and the user's request:
**Update an existing scenario** when:
- A new tool extends an existing workflow (e.g. adding `copy-form` to a CRUD scenario)
- The new tool is a natural addition to an existing test's prompt
**Create a new scenario** when:
- The new tool represents a different capability (e.g. search, hierarchy, import/export)
- Adding it to an existing scenario would make the prompt too complex
- No existing scenario covers the tool's domain
### Step 3: Update Existing Scenario (if applicable)
If updating an existing eval test:
1. Add the new tool to the `COLLECTION_TOOLS` array
2. Add the tool to `requiredTools` if it must be called
3. Update the prompt to include steps that exercise the new tool
4. Keep the prompt clear and sequential — don't make it overly long
**Build and run:**
```bash
npm run build
```
```bash
npm run test:evals -- --testPathPattern="{collection}"
```
If the test fails, iterate on the prompt. Common fixes:
- Make instructions more explicit
- Add a search/list step to find IDs dynamically
- Simplify the workflow if the prompt is too complex
### Step 4: Create New Scenario (if applicable)
Use the `eval-test-creator` agent to create `tests/evals/{collection}-{workflow}.test.ts`. The agent owns the test template and patterns.
**Build and run:**
```bash
npm run build
```
```bash
npm run test:evals -- --testPathPattern="{collection}-{workflow}"
```
Iterate on the prompt until the test passes reliably.
### Step 5: Report
Report what was done:
- Whether an existing scenario was updated or a new one was created
- File path
- Tools covered
- Test passing status
- Any prompt iterations needed
Related in AI Agents
skill-development
IncludedComprehensive meta-skill for creating, managing, validating, auditing, and distributing Claude Code skills and slash commands (unified in v2.1.3+). Provides skill templates, creation workflows, validation patterns, audit checklists, naming conventions, YAML frontmatter guidance, progressive disclosure examples, and best practices lookup. Use when creating new skills, validating existing skills, auditing skill quality, understanding skill architecture, needing skill templates, learning about YAML frontmatter requirements, progressive disclosure patterns, tool restrictions (allowed-tools), skill composition, skill naming conventions, troubleshooting skill activation issues, creating custom slash commands, configuring command frontmatter, using command arguments ($ARGUMENTS, $1, $2), bash execution in commands, file references in commands, command namespacing, plugin commands, MCP slash commands, Skill tool configuration, or deciding between skills vs slash commands. Delegates to docs-management skill for official documentation.
reprompter
IncludedTransform messy prompts into well-structured, effective prompts — single or multi-agent. Use when: "reprompt", "reprompt this", "clean up this prompt", "structure my prompt", rough text needing XML tags and best practices, "reprompter teams", "repromptception", "run with quality", "smart run", "smart agents", multi-agent tasks, audits, parallel work, anything going to agent teams. Don't use when: simple Q&A, pure chat, immediate execution-only tasks. See "Don't Use When" section for details. Outputs: Structured XML/Markdown prompt, quality score (before/after), optional team brief + per-agent sub-prompts, agent team output files. Success criteria: Single mode quality score ≥ 7/10; Repromptception per-agent prompt quality score 8+/10; all required sections present, actionable and specific.
adaptive-compaction
IncludedAdaptive add-on policy and recovery layer that decides WHEN to compact, prune, snapshot, or fork -- replacing fixed-percent auto-compaction across Claude Code, Codex, and MCP-capable hosts. Trigger on auto-compact timing or damage: "when should I compact", "is it safe to compact now or start a fresh session", "auto-compact fires too early/mid-task", "switching to an unrelated task but the window still has space", "context rot", "answers get worse the longer the session runs", "the agent forgot the plan or my decisions after it summarized", "add a layer on top that manages context without changing the agent", raising autoCompactWindow to give the policy room, or installing/tuning a cross-tool compaction policy or PreCompact hook -- even when "compaction" is never said but the problem is context-window pressure or post-summarization memory loss. Do NOT use to summarize a conversation, build RAG, write a summarization prompt (decides WHEN not HOW), or answer max-context-length trivia.
agent-skill-creator
IncludedCreate cross-platform agent skills from workflow descriptions. Activates when users ask to create an agent, automate a repetitive workflow, create a custom skill, or need advanced agent creation. Triggers on phrases like create agent for, automate workflow, create skill for, every day I have to, daily I need to, turn process into agent, need to automate, create a cross-platform skill, validate this skill, export this skill, migrate this skill. Supports single skills, multi-agent suites, transcript processing, template-based creation, interactive configuration, cross-platform export, and spec validation.
llm-wiki
IncludedUse when building or maintaining a persistent personal knowledge base (second brain) in Obsidian where an LLM incrementally ingests sources, updates entity/concept pages, maintains cross-references, and keeps a synthesis current. Triggers include "second brain", "Obsidian wiki", "personal knowledge management", "ingest this paper/article/book", "build a research wiki", "compound knowledge", "Memex", or whenever the user wants knowledge to accumulate across sessions instead of being re-derived by RAG on every query.
skill-master
IncludedAgent Skills authoring, evaluation, and optimization. Create, edit, validate, benchmark, and improve skills following the agentskills.io specification. Use when designing SKILL.md files, structuring skill folders (references, scripts, assets), ingesting external documentation into skills, running trigger evals, benchmarking skill quality, optimizing descriptions, or performing blind A/B comparisons. Keywords: agentskills.io, SKILL.md, skill authoring, eval, benchmark, trigger optimization.