@jhubbardsf/slack-mcp-server
v1.3.1
Published
Fork of slack-mcp-server with a resilient users-cache walk. Upstream: https://github.com/korotovsky/slack-mcp-server
Maintainers
Readme
@jhubbardsf/slack-mcp-server
Fork of slack-mcp-server by Dmitrii Korotovskii, MIT licensed.
Adds resilience to the startup users.list walk, which on very large
workspaces takes many minutes and previously discarded the entire enumeration on
a single transient failure:
- retries transient failures (5xx, 429) instead of aborting the walk
- checkpoints the cache every N pages, so a failure keeps prior progress
- never persists a partial roster over a cache that is already complete
Upstream PR: https://github.com/korotovsky/slack-mcp-server/pull/338
Usage
npx @jhubbardsf/slack-mcp-server --transport stdioConfiguration is unchanged from upstream. Two additions:
| Variable | Default | Purpose |
|---|---|---|
| SLACK_MCP_USERS_CHECKPOINT_PAGES | 10 | Pages between cache checkpoints; 0 disables |
| SLACK_MCP_USERS_MAX_RETRIES | 6 | Retry attempts per page, backoff capped at 60s |
Built from commit 807e6f6.
