markusdowne/agentmail
Give AI agents their own email inboxes using the AgentMail API. Use when building email agents, sending/receiving emails programmatically, managing inboxes, handling attachments, organizing with labels, creating drafts for human approval, or setting up real-time notifications via webhooks/websockets. Supports multi-tenant isolation with pods.
95
93%
Does it follow best practices?
Impact
100%
1.07xAverage score across 4 eval scenarios
Advisory
Suggest reviewing before use
Quality
Discovery
100%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is a strong skill description that clearly identifies the specific API (AgentMail), lists concrete capabilities, and provides an explicit 'Use when...' clause with diverse and natural trigger terms. It uses proper third-person voice throughout and is concise yet comprehensive, making it easy for Claude to distinguish from other skills and select appropriately.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions: giving AI agents email inboxes, sending/receiving emails programmatically, managing inboxes, handling attachments, organizing with labels, creating drafts for human approval, setting up real-time notifications via webhooks/websockets, and multi-tenant isolation with pods. | 3 / 3 |
Completeness | Clearly answers both 'what' (give AI agents email inboxes via AgentMail API, with specific capabilities listed) and 'when' (explicit 'Use when...' clause covering building email agents, sending/receiving emails, managing inboxes, handling attachments, etc.). | 3 / 3 |
Trigger Term Quality | Excellent coverage of natural trigger terms: 'email agents', 'sending/receiving emails', 'inboxes', 'attachments', 'labels', 'drafts', 'webhooks', 'websockets', 'AgentMail', 'programmatically', 'multi-tenant'. These cover a wide range of terms a user building email-related agent functionality would naturally use. | 3 / 3 |
Distinctiveness Conflict Risk | Highly distinctive — the combination of 'AgentMail API', 'AI agents', 'email inboxes', and specific features like 'pods' and 'drafts for human approval' creates a clear niche that is unlikely to conflict with generic email or messaging skills. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
85%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a strong, comprehensive API skill with excellent actionability — nearly every feature is demonstrated with executable code in both TypeScript and Python. The workflow clarity is solid with explicit validation checkpoints and error handling patterns. The main weakness is some verbosity, particularly the extended API key alias mapping discussion and the tangential Ajv validation section, which dilute the core skill content.
Suggestions
Significantly trim the API key alias mapping section — a 2-line fallback example with a brief comment would suffice instead of the multi-paragraph explanation with proof steps.
Consider moving the 'Strict JSON Validation For Read-Only Exports' section to a reference file, as Ajv configuration is tangential to the core AgentMail SDK usage.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is mostly efficient with good code examples, but the API key alias mapping section (Setup + 'Explicit alias mapping after proof') is overly verbose for an edge case. The 'Strict JSON Validation For Read-Only Exports' section with Ajv details feels tangential to the core AgentMail skill. The telemetry note at the top is unexplained noise. | 2 / 3 |
Actionability | Nearly every section provides fully executable code in both TypeScript and Python. Examples are copy-paste ready with real method signatures, parameter names, and return value usage. The code covers all major operations: inboxes, messages, threads, attachments, drafts, pods, batch operations, and error handling. | 3 / 3 |
Workflow Clarity | The Multi-Step Workflow Example section has explicit validation checkpoints between steps (verify inbox.inboxId before sending, verify sent.messageId before fetching). The Batch Operations section includes proper error isolation with try/catch per item and retry guidance. The JSON validation section includes a 4-step sequence with read-back verification. | 3 / 3 |
Progressive Disclosure | The skill is well-organized with clear section headers for each feature area. Advanced topics (webhooks, websockets) are appropriately referenced as one-level-deep links to reference files. The content flows from setup → basic operations → advanced patterns → safety considerations in a logical progression. | 3 / 3 |
Total | 11 / 12 Passed |
Validation
90%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 10 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
skill_md_line_count | SKILL.md is long (530 lines); consider splitting into references/ and linking | Warning |
Total | 10 / 11 Passed | |
Reviewed
Table of Contents