{"wiki":{"id":2851,"slug":"tr-ar-01","entity_type":"droid","entity_id":"tr-ar-01","title":"Tactical Reviewer","prose_md":"## About\n\nTactical Reviewer checks tactical plans for policy compliance, suitable owner assignments, missing steps, and realistic estimates. It gives [[agt_025]] a dependable baseline review, leaving harder judgment calls to its owner.\n\nWhen triggered on demand, it receives an execution ID and agent ID, applies its review rubric, and returns a pass/fail score with any problems found. It retries failures up to three times, then escalates unresolved reviews to [[agt_001]]. It cannot change, approve, or authorize plans, and it does not assess technical feasibility.\n\nIts exact name is unique, but it belongs to the reusable reviewer family, which has 31 droids run by several agents. It is pure software and does not call Claude.","namesake_json":"{\"engine\": \"software\", \"model\": null, \"claude_purpose\": null, \"family\": \"reviewer\", \"family_size\": 31, \"is_unique\": true}","profile_pic_path":"","source_hash":"1ed318513324ba5384abe18794ff6b6497edcbf808a751bfb74902a100a5a8f3","status":"active","last_generated_at":"2026-08-04 01:39:50","created_at":"2026-08-04 01:39:50","updated_at":"2026-08-04 01:39:50"},"facts":{"id":"tr-ar-01","name":"Tactical Reviewer","machine_name":"TR-AR-01","owner":"agt_025","function":"Reviews tactical plan accuracy and policy compliance","process_type":"triggered","schedule":"On demand","process_name":"tactical_reviewer","avatar_color":"#06b6d4","created_for":"Sage must review tactical plans to ensure they are sound, properly assigned, and compliant before execution begins. TR-AR-01 handles the technical accuracy check—verifying that plans meet baseline quality and policy standards so Sage can focus on strategic review.","purpose":"The droid evaluates each tactical plan against a concrete rubric: policy compliance, correct owner assignments, step completeness, and realistic estimates. This automated baseline check lets Sage spend time on harder judgment calls instead of routine soundness checks.","duties":"- Checks that policies cited in the plan actually exist and apply to the work\n- Verifies assigned agents are capable of their assigned steps\n- Looks for gaps or missing steps in the plan checklist\n- Judges whether time and resource estimates are reasonable\n- Retries failed reviews up to three times before giving up\n- Returns a pass/fail score and lists any problems found","constraints":"Reviews only—does not modify, approve, or authorize plans. Cannot fix problems on its own; escalates to Albus if reviews fail after retries. Requires an execution ID and agent ID as input; cannot discover or infer these. Focuses on structure and policy fit, not the correctness of the plan's logic or technical feasibility.","status":"running","gender":"male","archetype":"reviewer","building_block":"plan_step","service_override":null,"enabled":1,"task_key":"initiatives.review_verdict","model_free":0},"page":{"type":"droid","page_class":"main","class_label":"Main page","kind_label":"Droid","kind_gloss":"","listed":true}}