Repository navigation
Rewrite the observability investigate and alert-investigation guides around the current MCP tools - #241
Merged
Conversation
… tools investigate gets a first-five-minutes sequence (scope and onset, chronic or noise, get-changes, breakdown against a baseline with compareTo, evidence, blast radius) and a sampling warning for percentiles. alert-investigation starts from get-alerts and get-alert-firing-history instead of injected alert variables. Companions use the stable search-* tool names. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Generated-by: termag
Vadman97
marked this pull request as ready for review
November 5, 2026 01:27
![cursor[bot]](https://avatars.githubusercontent.com/in/1210556?s=60&v=4){kind=small}
Contributor
Author
|
bugbot run |
Skill eval results
Only suites whose source actually changed since their last recorded score were re-run. Soft-failing while we stabilise the baseline. |
Firing-history summary (chronic, episodes) and get-changes alert_episode rows in alert-investigation; get-trace self_ms, critical_path and top_self_time; project-wide search-flag-evaluations and the FLAG FILTERS keys; the low_sample warning; get-error-group include and _pivots; sort spike; and the ingest block with completeness_note. Each names a fallback for servers that predate it. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Generated-by: termag
Conflicts were main's query-* to search-* tool rename (#231) in files this branch already rewrote with the search-* names and camelCase parameters; kept the branch versions. CHANGELOG keeps both sides' entries. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Generated-by: termag
Contributor
Author
|
bugbot run |
…ja9ob4ngjy83vmxi7472 Generated-by: termag # Conflicts: # CHANGELOG.md
Contributor
Author
|
bugbot run |
There was a problem hiding this comment.
✅ Bugbot reviewed your changes and found no new issues!
Comment @cursor review or bugbot run to trigger another review on this PR
Reviewed by Cursor Bugbot for commit 46e1e7e. Configure here.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Demo
Replay of im-live incidents 1 and 3 on catfood following the new guides: demo.md (
team-o11y:artifacts/im-guides/demo.md). INC-298 took 12 calls instead of 14, and the version was ruled out at call 7 instead of 13. Alert 166 took 6 calls instead of 13, with the chronic-noise verdict at call 3.Rewrites
observability/investigate(0.3.0) around a "first five minutes" sequence: onset, chronic check,get-changes, a breakdown against a baseline withcompareTo, evidence, and blast radius, plus a warning about sampled percentiles. Rewritesobservability/alert-investigation(0.2.0) to start fromget-alertsandget-alert-firing-historyinstead of injected alert variables.The guides also use the fields from launchdarkly/mcp#449, each with a fallback for older servers. Merge this after #449 deploys. It conflicts with #231 in the observability files; resolve those by taking this PR's version, which already uses the stable
search-*names.Asked by
vkorolikthroughtermag ask; follow up withtermag ask … --thread cmw2uja9ob4ngjy83vmxi7472.🤖 Generated with Claude Code
Opened from termag
Note
Overview
investigate0.3.0 replaces the old cross-product workflow with a ordered “first five minutes” playbook: scope/onset (including ingest completeness), chronic vs regression checks withcompareTo,get-changes, dimensional over-representation (host, version, route, flag variation), then evidence and blast radius viasearch-service-map,search-log-patterns,get-trace,get-error-group, etc. It drops the old “50-entry limit first” framing in favor of aggregate-then-page guidance, adds sampling/percentile (low_sample) warnings, structured output, and documents launchdarkly/mcp#449 response fields with fallbacks for older servers.alert-investigation0.2.0 no longer assumes alert variables injected on the first turn. Agents now start fromget-alertsandget-alert-firing-history(replay via bundledsearch-aggregationsargs, chronic/episodes fromsummary, optionalget-investigation), add a Verdict section (noise/chronic/new), and delegate root-cause steps to the shared investigate sequence. Per-product companions (logs, traces, errors, metrics, sessions) are aligned to replay, baseline comparison, and the new tools.skills.jsondescriptions and versions are updated; CHANGELOG records the release notes. Intended to ship after MCP #449 deploys.Reviewed by Cursor Bugbot for commit 46e1e7e. Bugbot is set up for automated code reviews on this repo. Configure here.