feat: DISABLE_SKILLS experiment toggle + tool arg schema improvements#444
feat: DISABLE_SKILLS experiment toggle + tool arg schema improvements#444isiddharthsingh wants to merge 3 commits into
Conversation
|
Important Review skippedDraft detected. Please check the settings in the CodeRabbit UI or the ⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: ASSERTIVE Plan: Pro Run ID: You can disable this status message by setting the Use the checkbox below for a quick retry:
✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
|



Summary
DISABLE_SKILLSenv-var toggle that bypasses integration skill markdown and unregisters theload_skillLLM tool, so the agent can be A/B tested relying purely on tool descriptions.CloudExecArgs,IacToolArgs,TerminalExecArgs,AnalyzeZipArgs,OnPremKubectlArgs) so the LLM sees enum hints and per-field descriptions in the tool JSON schema. Adds alist_onprem_clustersdiscovery tool so the agent can find cluster IDs without skill guidance.interactive_load_skillsegment +SkillRegistryintegration methods short-circuit when the flag is on; identity, security, behavioral_rules, RCA scaffolding, and core prompts keep loading.docker-compose.yaml,docker-compose.prod-local.yml,docker-compose.airtight.yml,deploy/helm/aurora/values.yaml, and.env.example.Test plan
tests/experiment/test_disable_skills.py)list_onprem_clustersreturns thestatus:ok|errorenvelope; verified the error path for unresolvable usersload_skilltool is unregistered when the flag is on