From flagos-skills
Orchestrates end-to-end TLE development: writing/optimizing high-performance kernels, implementing API/verifier/lowering features, debugging correctness/performance issues with reproducible validation.
npx claudepluginhub flagos-ai/skills --plugin flagos-skillsThis skill uses the workspace's default tool permissions.
Use this skill to execute TLE work end-to-end:
Generates and optimizes GPU kernel operators for Python/Triton, FlagGems, and vLLM repositories. Auto-detects repo type, dispatches sub-skills for MCP-based optimization, platform specialization, and feedback.
Implements coding tasks via Test-Reinforced Development: separates Overseer test-writing from Executor implementation with Coordinator enforcement and codex cross-reviews to prevent cheating.
Implements features using parallel subagents with scope control, reflection, and MCP servers for memory/context. Activates on implement/build/create requests in JS/TS projects.
Share bugs, ideas, or general feedback.
Use this skill to execute TLE work end-to-end: intake -> implementation -> validation -> artifacts -> merge decision.
Use for:
references/.Start every task with:
Goal:
Non-goal:
Acceptance:
Impact scope (optional):
references/tle-sources.mdreferences/workflow-templates.mdreferences/tle-sources.md as the technical source of truth for:
references/workflow-templates.md as the source of truth for:
// begin flagtree tle and // end flagtree tle.third_party/tle.references/tle-sources.mdreferences/workflow-templates.md