next-steps — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited next-steps (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.
Reads the current conversation (and optional connected context: Linear, Slack thread, Notion doc, git/PR state) and produces one consolidated checklist of next steps grouped by owner, ranked by priority, with explicit dependencies and open questions. Designed to be pasted into a Slack thread, a Linear comment, or a Notion update so everyone sees the same picture.
The point isn't to enumerate every possible task — it's to surface what unblocks the next move forward, so multiple stakeholders can self-serve.
consolidate.
on next items", "summarise what each person owes"*.
Do NOT use this for: timesheet recaps (use /workday-summary), retrospective status posts (use /project-updates), or commit-by-commit breakdowns.
Required:
has this in context.
Optional (from the scope-hint arg or follow-up questions):
ENG-1234) — fetch via mcp__linear-server__get_issue to anchor scope and check current state/assignee.#822) — fetch via gh pr view for state, reviewers, blocking checks.If the conversation already contains these references (it usually does), pull them automatically. Only ask the user when scope is genuinely ambiguous.
Identify, in one sentence, what project / piece of work this is for. Pull it from the conversation. If multiple candidates exist, ask the user which one.
List every named person and role that has appeared in the conversation or in referenced tickets/threads. Mark each as one of:
Don't invent stakeholders. Use names actually mentioned.
Walk the conversation and any fetched docs. For each item, capture:
P0 (blocks demo / today), P1 (this week), P2 (next pass)Apply ranking ruthlessly: an item that unblocks another stakeholder outranks an item that's purely internal polish.
Anything that's a decision needed — not a task — goes into a separate Open questions section, attributed to the decider. These are the things that, once answered, generate concrete tasks. Don't merge them with action items.
# Next steps — {project / scope}
## {Owner name or @handle}
- [ ] **P0** — {action}. _{context, source}_
- [ ] **P1** — {action}. _Blocked on: {dep}_
## {Next owner}
- [ ] **P1** — {action}.
## Open questions
- **{Decider}**: {question}. _Why it matters: {one clause}_
## Out of scope (parking lot)
- {item} — revisit when {trigger}Hard rules:
name should see all their items in one block.
flag it and ask which to drop.
"Vineet on Apr 27", "see Notion guide §3").
Show the rendered checklist in chat. Ask:
Iterate once, then offer:
pbcopy) for Slack/Notion pasteapproval — never auto-post)
~/Desktop/next-steps-{slug}.md) for sharingoptions for un-merge functionality"*.
conversation, the IA placement question is open"*.
outcomes, no engineering minutiae.
current scope and a real stakeholder.
default to P1 and flag "priority unconfirmed".
should be omitted, not filled with "none".
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.