ClawTrust
8.7
/ 10
1 evaluations
2.2k Downloads
Overview
Provide a deterministic, JSON-first CLI interface to the Notion API for searching, reading/exporting, creating/importing, appending, moving, and triaging Notion pages and database-backed notes, optimized for use by automation agents.
Key Advantages
1.Single, well-defined CLI entrypoint (notionctl.mjs) with predictable arguments, ideal for OpenClaw-style automation and allowlists.
2.JSON-first design and deterministic behavior, reducing error rates around headers, pagination, rate limits, and retries compared with ad-hoc API calls.
3.Broad coverage of common Notion workflows: search (pages and data sources), export to Markdown, create from Markdown (under pages or databases), append content, move pages, and inbox triage.
4.Built-in operational guidance for safe bulk edits, including dry-run / no-apply modes, scope limiting, and explicit rules-driven triage flows.
5.Clear handling of Notion API specifics (Notion-Version, data_source_id vs database_id, rate limiting, Retry-After semantics) and authentication fallbacks via environment variables and config file.
Use Cases
- Automated capture of chat outputs or ideas into a Notion inbox (page or database) as Markdown notes with relevant metadata and provenance.
- Inbox triage workflows where an agent periodically classifies and moves pages from an inbox page into the correct pages or databases using rules files and dry-run verification.
- Searching across Notion pages or databases for relevant content before answering user questions or deciding where to write new information.
- Exporting Notion pages as Markdown for downstream processing, summarization, or backup, then optionally appending updates back into those pages.
- Programmatically reorganizing a workspace by moving pages between parent pages or into specific databases (data sources) while respecting Notion API constraints.
Evaluation Scores
8.7
/ 10
Reliability
8.8
Functionality
8.2
Usability
8.5
Safety
9.3
Performance
8.3
Compatibility
9.0
Based on 1 evaluation · Latest: 3/19/2026
Download Trend
Loading...
Evaluation History (1)
8.7/103/19/2026▼
OS: darwin-x64LLM: stepfun/step-3.5-flash
**Judgement:** This skill is a strong, automation-friendly interface to the Notion API, well-suited for agents that need reliable, repeatable Notion workflows (search, export, create, append, move, triage) rather than one-off API calls.
**Strengths:**
- Single, deterministic CLI entrypoint that aligns well with OpenClaw’s expectations (predictable args, JSON orientation).
- Good coverage of common Notion operations including inbox capture and triage, Markdown-based read/write, and moving pages into databases using the correct `data_source_id` semantics.
- Thoughtful operational guidance around rate limits, retries, pagination, and error handling, plus troubleshooting notes for 401/403/404/429 and validation errors.
- Strong safety posture: explicit instruction to treat Notion content as untrusted input, prefer read-then-decide-then-write patterns, and use dry-run/limits for bulk edits.
**Risks & Limitations:**
- Misconfigured bulk operations or triage rules can still cause large-scale moves or edits; agents must respect `--dry-run`/`--apply` and `--limit` recommendations and be conservative by default.
- Functionality appears focused on Markdown-centric workflows and high-level operations; more advanced database querying or fine-grained property manipulation may be limited or require additional custom commands not evident in the snippet.
- Requires correct environment configuration (`NOTION_API_KEY` or fallbacks); missing or under-scoped tokens will manifest as 401/403/404 errors that agents must interpret correctly.
**Recommended Scenarios:**
- Agents capturing conversation outputs into a Notion inbox and then periodically triaging those notes into structured databases or project areas.
- Systems that need to read and update Notion pages in Markdown form (status updates, running logs, appended summaries) with low risk of API misuse.
- Workspace reorganization or routing tasks where the agent searches, moves, and classifies pages according to stable rules, especially when bulk operations are guarded by dry-runs and limits.
Comments (0)
No comments yet. Be the first!