Sprint Retro Facilitator — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited Sprint Retro Facilitator (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Most retros produce the same three actions every sprint and close nothing. This skill forces specificity, surfaces the issues people are actually thinking, and ends with someone's name on every action.
Review the previous retro's action items first. Open with their status: done, in progress, or dropped. If more than half were dropped, say so plainly — this is itself a retro topic. Set a timer and a note-taker. Aim for 60 minutes maximum; 45 is better.
Use a silent brainstorm before discussion — individuals write items on sticky notes or a shared board without seeing each other's input. This prevents the most vocal person from anchoring the group. Prompt with three columns: 'What slowed us down?', 'What went well that we should protect?', 'What are we not saying out loud?' The third column is the most valuable and most often skipped.
Give each person 3 votes. Cluster near-duplicate items. The top 2-3 clusters by votes become the discussion topics. This respects the group's actual priorities, not the facilitator's.
For each topic: state the observation as a system problem, not a person problem. Ask 'what made this likely to happen?' before 'what should we do?' Root-cause one level deeper than the surface complaint — if the answer is 'we need better communication,' that is not a root cause. Push for the structural or process condition that produced the failure.
Every action item needs three fields: what exactly will change, who owns it (one name, not 'the team'), and when it will be done or reviewed. Cap actions at 3 per retro. More than 3 almost never close. If there are more, vote on the top 3 and park the rest in a visible backlog.
If the same item appears in three consecutive retros with no progress, escalate it — it is either blocked by something outside the team's control or no one actually owns it. Name both possibilities explicitly.
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.