eval: rh-sre-fleet-inventory#22
Open
GuyZivRH wants to merge 20 commits into
Open
Conversation
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Strip methodology from instruction, add tests for stale <7d check-in heuristic, Vulnerable/Patched/Not Affected status strings, system UUID tracking, and EOL RHEL compliance flagging. Co-authored-by: Cursor <cursoragent@cursor.com>
Replace broad keyword matching with case-sensitive checks for exact fields: Vulnerable/Patched/Not Affected status strings, stale, last_seen, remediation_available, get_cve_systems, display_name, fqdn. Co-authored-by: Cursor <cursoragent@cursor.com>
CVE- appears in any CVE report trivially. Now requires the exact tool name get_cve_systems for meaningful discrimination. Co-authored-by: Cursor <cursoragent@cursor.com>
…wledge Tests now check for: case-sensitive status strings (Vulnerable/Patched/ Not Affected), last_seen staleness, remediation_available flag, display_name/fqdn identifiers, get_host_details/get_cve_systems tools. Co-authored-by: Cursor <cursoragent@cursor.com>
- Add CLAUDE.md as treatment-only system prompt - Move docs from supportive/docs/ to docs/ at submission root - Nest skills under skills/fleet-inventory/ with sibling mcp-lightspeed-validator - Fix all doc reference paths to /docs/ for container resolution - Add skill usage hint to instruction.md - Sync enhanced mock-lightspeed-mcp.py with pagination support Co-authored-by: Cursor <cursoragent@cursor.com>
Wire up the .ai-index/ semantic index so the skilled agent discovers the docs library via CLAUDE.md system prompt. Also deduplicate prior accidental repetitions. Co-authored-by: Cursor <cursoragent@cursor.com>
Subagents were refusing to call MCP tools (e.g. inventory__list_hosts) because CLAUDE.md's skill-first rule was interpreted too literally, causing deadlock loops and timeouts during fleet-inventory evaluation. Co-authored-by: Cursor <cursoragent@cursor.com>
Tests checked for exact snake_case API field names (display_name, fqdn, last_seen, remediation_available) but the agent writes readable column headers. Accept both forms. Also handle empty LLM responses in judge. Co-authored-by: Cursor <cursoragent@cursor.com>
Replace vague conceptual checks with get_host_details tool, lightspeed validator prereq, fleet size (63 systems), specific CVE references (CVE-2024-12345), environment breakdown, and remediation transition. Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
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
1 participant
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.
A/B evaluation for rh-sre fleet inventory skill.
Made with Cursor