You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Accept both issue ID formats in issue: and issue.id: filters #130940
Event search has two filters for the same thing, and each accepts only one issue ID format:
issue: accepts only qualified short IDs (PROJECT-123). A numeric group ID such as issue:4567 fails the short-ID lookup and returns a 400: Issue IDs do not exist: ['4567']. This happens in the Discover/errors dataset (_issue_filter_converter in src/sentry/search/events/datasets/discover.py) and in the EAP occurrences dataset (_init_issue_short_id_cache in src/sentry/search/eap/resolver.py).
issue.id: is a numeric key in the search parser (default_config in src/sentry/api/event_search.py), so issue.id:PROJECT-123 fails to parse with Invalid number.
Other APIs return numeric issue IDs, so API and MCP clients that pass them to issue: get a 400. The Sentry MCP server now rewrites numeric issue: filters to issue.id: on the client to work around this (fix(mcp): Accept numeric issue IDs in error searches toolkit#2204).
Proposal from @dcramer: issue: should accept every format, and issue.id: should accept short IDs. There is no value in keeping the two split.
Activity
linear-code commentedon Nov 12, 2026
ENG-9037
fix(search): Keep issue.id 0 and resolve uncached short IDs in EAP is…
fix(search): Scope issue short-ID lookups to the query's projects