Trace plan claims through code and report execution mismatches.
npx skills add https://github.com/exerias21/brainstorm-toolkit --skill flowsim
Please help me install this Agent Skill. Skill: flowsim Source: https://github.com/exerias21/brainstorm-toolkit/tree/main/skills/flowsim Command: npx skills add https://github.com/exerias21/brainstorm-toolkit --skill flowsim
Trace claims in plans or tasks through the actual source code and surface mismatches in the resulting execution narrative. It enables structured code-review passes that reveal where the implementation diverges from the plan, improving the reliability of sdlc cycles.
Run flowsim with a plan file or task reference to begin tracing the end-to-end flow through the codebase.
High-intent search queries and answers about installing and using this skill.
You trace plan-vs-code mismatches by consuming a plan file or task reference and mapping claimed flows to concrete code anchors. The trace reports mismatches where actual code execution diverges from the intended implementation narrative.
Plan-vs-code validation in SDLC workflows is the process of tracing claimed flows through actual source code after feature implementation. It ensures implementation alignment by surfacing mismatches between the planned execution narrative and the written code anchors.
Control tracing depth and scope using the --max-hops, --focus, and --force options. These parameters adjust how deeply the trace traverses the codebase and allow targeted analysis of specific execution flows during ad hoc investigations.
Yes, flowsim supports re-tracing and caching to reuse prior traces when code hasn't changed. This saves time across fix loops by avoiding redundant tracing operations during repeated plan-to-code validation passes.
Yes, plan-to-code tracing correlates with eval results and tests to validate flow correctness. This integration ensures that the traced execution narrative aligns with actual test outcomes and evaluation metrics across the codebase.
Run plan-vs-code validation after feature implementation within SDLC workflows to ensure alignment. It performs a structured code-review pass that reveals where the implementation diverges from the plan, improving overall cycle reliability.
deandumGrep the repository to identify related code and write a discovery.md handoff plan.
majiayu000Route user requests to registered skill modules and execute them in context.
Imad-OuteAutomate code reviews for correctness, security, and testability in SDLC phase 6.
yun520-1Analyze codebases for architectural patterns, code quality, and DDD elements.
