Production-grade Slack MCP server with OAuth, observability, safe file uploads, and a CLI.
15 tools with consistent verb-noun naming (list_, get_, create_, send_, etc.) and complete input schemas. All tools have descriptions (avg ~60 chars, baseline 194). Parameters are typed and described. However, descriptions are notably brief, most under 100 chars vs. baseline 194, limiting LLM context for tool selection. Output schemas are not documented in the source. Error handling is present but generic. No tool annotations (readOnlyHint, destructiveHint, idempotentHint) despite clear risk levels. Missing dependency hints and recovery guidance in descriptions.
Add a reaction emoji to a Slack message
Archive a Slack channel
Create a new Slack channel
Delete a Slack message
Get detailed information about a specific Slack channel
Get detailed information about a specific file in Slack
Retrieve messages from a Slack channel
Retrieve messages from a specific thread in a Slack channel
Descriptions are too brief (avg ~60 chars vs. baseline 194). Most lack WHEN to use, prerequisites, or recovery hints. E.g., 'List all channels in a Slack workspace' omits: when to call this, what structure it reveals, or how to filter results.
No output schemas documented. LLMs cannot plan downstream calls or extract required fields. E.g., does list_channels return channel_id, name, topic, member_count? Unknown.
No tool annotations (readOnlyHint, destructiveHint, idempotentHint) despite clear risk levels. delete_message, archive_channel, and update_message are marked DESTRUCTIVE/REVERSIBLE/WRITE but lack protocol-level hints to guide agent caution.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 50 | <=2025-11-25 | v2 |
Get detailed information about a specific Slack user
List all channels in a Slack workspace
List files in a Slack channel
List all users in a Slack workspace
Send a message to a Slack channel
Update an existing Slack message
Upload a file to a Slack channel
Error handling is generic. No recovery guidance in tool descriptions. E.g., 'Channel not found' does not suggest 'Try list_channels() to find the correct ID' or offer available alternatives.
Pagination parameters (limit, oldest) lack constraints. No min/max documented. LLMs may pass absurd values (limit=999999) causing API timeouts or excessive token consumption.