fix: document cypher patterns and forbid direct JSON file access#70
Merged
fix: document cypher patterns and forbid direct JSON file access#70
Conversation
Add explicit cypher query examples and a CRITICAL instruction to use execute_cypher for all result retrieval. Models were bypassing cypher and reading /tmp/patent-search-*.json with python3, causing redundant file reads and wasted time. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Replace runner.sh and tools/*.sh with skill-bench CLI native TOML format. Move test cases from agents/skill-bench/cases/ to tests/. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Resolve conflicts: - Update test files to use 'tool-param' instead of 'tool-use' - Keep HEAD versions of test prompts (Google/Toyota) for consistency - Accept main branch updates to Cargo.lock, models.rs, and other core files Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
execute_cypherquery patterns topatent-searchandpatent-assignee-checkskillspatent-searchpatent-assignee-checkin line withpatent-fetchwhich already had this guardContext
Models were bypassing
execute_cypherand reading/tmp/patent-search-*.jsonfilesdirectly with
cat | python3, causing 9+ redundant file reads per test run.This reduced
targeting/functional-with-spectest from 323s to 146s (55% faster).Test plan
npx prettier --writeandclaude plugin validate .passpatent-kittargeting test: JSON reads dropped from 9 to 0,execute_cypherusage increased🤖 Generated with Claude Code