{"id":"lead-blameless-retro-rewrite-statements","audience":"managers","category":"leadership-work","track":"leadership","level":"intermediate","technique":"tone","title":"Rewrite blame statements from a retro into blameless findings","free":false,"task":"Rewrite five blame-style comments from an incident retro into blameless findings about the process. Write exactly 5 numbered lines in the format: 1. Finding: ... Fix: ... Do not name any person, do not use the words careless, failed, forgot or blame, each line stays under 30 words, and every Fix starts with one of the verbs Add, Require, Route, Assign or Create. Use only what the comments say.","materialLabel":"Retro comments (attached to your prompt automatically)","material":"Incident: Nov 6 outage, 2 hours, checkout down.\n\n1. \"Raj forgot to run the database migration check before deploying.\"\n2. \"Priya was careless and approved the pull request without testing.\"\n3. \"Nobody told on-call that the deploy was happening, Sam should have posted in the channel.\"\n4. \"The alert fired at 2:10 pm but Li failed to see it for 25 minutes.\"\n5. \"Marco blamed the vendor, but the real issue is that he never read the vendor's status page.\"\n\nFacts from the timeline: the deploy checklist has no migration step; PRs have no required test evidence; there is no deploy announcement step; alerts go to a channel with 90 members; no one owns the vendor status page.","rules":[{"type":"numberedItems","count":5},{"type":"maxWords","max":170},{"type":"plainText"}],"mistakes":[{"title":"Naming people","why":"The AI keeps 'Raj' or 'Priya' in the finding.","fix":"Say: \"Don't name anyone; describe the process gap.\"","checks":["r1"]},{"title":"Judgmental words","why":"The AI keeps 'careless' or 'forgot'.","fix":"Say: \"Ban the words careless, failed, forgot, blame.\"","checks":["r1"]},{"title":"Vague fixes","why":"The AI says 'communicate better'.","fix":"Say: \"Each fix is a concrete process change from the facts.\"","checks":["r2"]},{"title":"Wrong format","why":"The AI writes a paragraph or merges findings.","fix":"Say: \"Format: 1. Finding: ... Fix: ...\"","checks":["r0","numberedItems"]},{"title":"Trusting the AI to count words","why":"AI is bad at counting. Ask for a total word limit and it often lands over it.","fix":"Cap each line: \"each line under 30 words.\"","checks":["maxWords"]},{"title":"Trusting the AI to count words","why":"AI is bad at counting. Ask for a total word limit and it often lands over it.","fix":"Cap each part instead, for example: \"each line under 12 words.\"","checks":["maxWords"]},{"title":"Adding headings and bold","why":"The AI writes a formatted report with headings and an action list.","fix":"Say: \"Plain text: only the 5 numbered lines.\"","checks":["plainText"]}],"promptHints":[{"label":"Shows the format","patterns":["finding:.*fix:","1\\.\\s*finding"]},{"label":"Bans names","patterns":["don'?t name","no names","anonymous"]},{"label":"Bans words","patterns":["careless","forgot","blame","ban"]},{"label":"Wants process fixes","patterns":["process","checklist","concrete"]},{"label":"Counts lines","patterns":["5 (numbered )?lines","five"]}],"hint":"Show the line format, ban the names and the judging words, and require concrete process fixes.","rubricLabels":["Each line follows \"N. Finding: ... Fix: ...\" with five lines matching the five comments in order","No line names Raj, Priya, Sam, Li or Marco, and none uses the words careless, failed, forgot or blame","Every numbered line contains the labels \"Finding:\" and \"Fix:\", and every Fix starts with Add, Require, Route, Assign or Create and is a process change drawn from the timeline facts"]}