Evaluate task outputs against Task.md Expected Results and generate evaluation.md.
npx skills add https://github.com/hs3180/disclaude --skill evaluator-hs3180
Please help me install this Agent Skill. Skill: evaluator Source: https://github.com/hs3180/disclaude/tree/main/skills/evaluator Command: npx skills add https://github.com/hs3180/disclaude --skill evaluator-hs3180
This Skill evaluates whether a task has been completed according to Task.md Expected Results, producing a structured evaluation and, when complete, a final result signal.
Execute an evaluation cycle by reviewing the executor's output and writing an evaluation.md; if all criteria are met, also write final_result.md.
High-intent search queries and answers about installing and using this skill.
To evaluate task outputs, you compare the executor's deliverables against the Task.md Expected Results and generate an evaluation.md file documenting the status and required next actions.
Final_result.md signals task completion by acting as a deterministic success marker. It is only generated when the evaluation confirms all Expected Results criteria are fully met.
You do not need to manually write evaluation reports from scratch. The evaluation process automatically reviews executor outputs against criteria and enforces strict output conventions for evaluation.md.
The best way to document next actions is to include deterministic guidance within the evaluation.md. This ensures auditable progress tracking and explicitly states what changes are required for the next iteration.
You cannot use this evaluation process without a Task.md file. The mechanism strictly depends on comparing executor outputs against the Expected Results defined within Task.md to generate valid assessments.
VjayRamDefine DSPy signatures and modules with deterministic evaluation across splits.
Chanw-researchMonitor running experiments across local and remote environments and generate status reports.
KarubiOhayoGenerates a review report and relay file with Go/No-Go recommendation for executor deliverables.
nahisahoReview experimental results for reproducibility, statistical validity, and methodological soundness.
