stakeholder-alignment — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited stakeholder-alignment (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.
Decisions fail more often in the alignment than in the logic. A good RFC or decision doc does two things: makes the best argument for the chosen path, and makes the roles of everyone affected explicit so no one is surprised.
| Artifact | When to use | Length |
|---|---|---|
| RFC / design doc | Technical decision affecting multiple teams; invites comment | 3-10 pages |
| Proposal | Bet on a direction that requires budget/headcount/scope | 1-3 pages |
| Pre-read | Input for a scheduled decision meeting | 1-2 pages |
| Decision doc | Record a decision that has been made (or is being made now) | 1 page |
| One-pager | Early alignment on an idea before full RFC | 1 page |
Use the smallest artifact that produces alignment. A 10-page RFC for a two-team change is over-investment; a Slack thread for a 10-team change is under-investment.
# [Title] — [state: Draft / Reviewing / Accepted / Rejected / Superseded]
## BLUF
[One-paragraph summary: the problem, the proposed direction, the ask.]
## Roles (DACI)
- **Driver:** [who writes + shepherds]
- **Approver:** [who makes the final call — name, not group]
- **Contributors:** [who provides input]
- **Informed:** [who needs to know]
## Context
[Shared understanding. What led here. Keep controversial claims for later sections.]
## Problem
[What we are trying to solve. Why now.]
## Options considered
- Option A: [name] — [one paragraph] — pros / cons / rough cost
- Option B: [name] — [same]
- Option C: do nothing — [same]
## Recommendation
[Chosen option + why, in one paragraph.]
## Trade-offs
[What we are giving up. What we are not solving. Second-order effects.]
## Open questions
- [ ] [Question needing input] — @person
- [ ] [Assumption needing validation] — @person
## Plan
[Milestones with dates. Dependencies on other teams.]
## Appendix
[Supporting material, benchmarks, prior discussions.]Variations: for short changes, collapse Options / Recommendation into one section. For contentious changes, expand Trade-offs with a pre-mortem.
Worked examples and short-form variants in references/rfc-design-doc-templates.md.
Both frameworks make roles explicit. Pick one and use it consistently across the org.
| Role | Meaning |
|---|---|
| Driver | Shepherds the decision. Writes the doc, runs the meeting. |
| Approver | Makes the final call. Usually one person. |
| Contributors | Provide input. Their concerns must be addressed; they do not have veto. |
| Informed | Need to know once the decision is made. |
| Role | Meaning |
|---|---|
| Recommend | Proposes the decision. |
| Agree | Formal sign-off required (legal, security, etc.). |
| Perform | Executes once decided. |
| Input | Consulted. Not a veto. |
| Decide | Makes the call. |
Rules:
A pre-read makes a 30-minute decision meeting take 10 minutes.
Structure:
Sent 24-48 hours before the meeting. If the meeting still takes 30 minutes, either the pre-read was bad or the decision was not yet decision-ready.
A decision doc captures a decision AFTER it has been made. Purpose: future reference, onboarding, accountability.
# Decision: [title]
**Date:** [YYYY-MM-DD]
**Decider:** [name]
**Status:** Accepted / Superseded by [link] / Reversed
## Context
[Why a decision was needed.]
## Decision
[What was decided, in one paragraph.]
## Alternatives considered
[Brief: option → reason not chosen.]
## Consequences
[What this means going forward. What we are giving up.]
## Follow-ups
- [ ] [Action] — @owner — [date]This is distinct from an ADR (Architecture Decision Record) which focuses on architecture specifically — see documentation-discipline for ADRs.
When alignment fails:
| Anti-Pattern | Problem | Solution |
|---|---|---|
| RFC as rationalization | Doc written to justify a decision already made; readers disengage | Write the RFC before the decision; include real alternatives |
| Group approver | "Engineering leadership" as approver; no one decides | One named person as Approver/Decider |
| Missing trade-offs section | Only upsides listed; readers suspect downsides not examined | Every proposal lists what it gives up |
| Options theater | Option A is the real proposal; B and C are strawmen | Include serious alternatives or omit the section entirely |
| Consensus-seeking | Trying to get everyone to agree on everything | Decider decides while contributors contribute; disagreement is logged |
| Pre-reads read in the meeting | Meeting time wasted re-reading the doc | Send 24-48 hours before; if not read, doc was too late or too long |
| Missing kill condition | No clause for what evidence would reverse the decision | Add "We will reconsider if [specific measurable outcome]" |
| Decider named after debate | Roles migrate to favor whoever speaks loudest | Publish roles before the debate |
| Contributors treated as vetoers | Individual contributors block progress | Contributors provide input; they do not have veto power |
| File | Contents |
|---|---|
references/rfc-design-doc-templates.md | RFC templates (short + long), variants per decision type, naming conventions |
references/daci-rapid-cheatsheet.md | Full DACI and RAPID specs, role-assignment rules, common mistakes |
references/alignment-playbooks.md | Pre-read checklist, escalation patterns, async alignment vs sync meeting trade-off |
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.