name: align-to-mission description: "Use when you want to check whether every phase in a plan serves the project's stated mission. Catches scope creep, tangential work, misaligned success metrics, and aggregate drift from the north star." argument-hint: "[plan-file-path] [--dispatch]"
Align to Mission
Check whether every phase and deliverable in a plan serves the project's north star. Plans accumulate scope naturally — individually-reasonable phases can drift from the mission in aggregate. This skill compares each phase against the mission and flags misalignment before resources are committed.
This is a plan-only lens. It reads the plan document and the project's mission source. It does not read the target codebase (that is /extension-check's job).
Arguments
Parse $ARGUMENTS at invocation:
- First positional arg: Path to the plan document. If omitted, prompt for it.
--dispatch: Non-interactive mode for fleet orchestration. Suppresses all interactive elements (no Plan Mode, no AskUserQuestion). Emits a single markdown document with YAML frontmatter perskills/_shared/contracts/lens-result-contract.md. Use this when invoked by/ironclador another orchestrator.
--dispatch Mode
When --dispatch is passed:
- Do NOT call
EnterPlanMode. The dispatcher owns the lifecycle. - Do NOT call
AskUserQuestion. No human is present. - Do NOT prompt for clarification. If the plan lacks phases or is too ambiguous to assess, emit
status: blockedwith a description of what is missing. - Execute the procedure below silently.
- Emit the structured markdown result as the FINAL output and stop. No text after the result document.
When --dispatch is NOT passed, follow the interactive procedure (see Interactive Mode below).
Procedure
Step 1: Read the Plan
Read the full plan document. Identify:
- Goal / Problem Statement — what the plan claims to solve.
- Phases — the breakdown of work. Each phase is assessed individually.
- Decision Forks — choices that may affect alignment.
- Validation Strategy — success criteria the plan defines for itself.
If the plan lacks identifiable phases, stop. In --dispatch mode, emit status: blocked. In interactive mode, tell the user the plan needs a phased breakdown before alignment analysis is meaningful.
Step 2: Locate the Mission
Find the project's north-star statement using this fallback hierarchy:
PROJECT_MISSION.mdin the repo root — the canonical source.README.mdmission/purpose section — extract the north-star statement.- GitHub repo description — the one-liner from the repo settings (if accessible).
CLAUDE.mdpurpose/description section — the repo's self-description for Claude Code.
Use the first source that yields a usable mission statement — a concrete statement of what the project exists to achieve, not a list of features or technical description.
If none of these yield a usable mission statement, do NOT block. Instead:
- Emit an
info-severity finding recommending the team create aPROJECT_MISSION.md. - Infer the mission from the plan's own Goal section and proceed with that as the working mission. Note explicitly that you are using the plan's goal as a proxy — this weakens the analysis because the plan is being compared against itself.
Step 3: Per-Phase Alignment Assessment
For each phase or deliverable in the plan, assess:
3a: Does This Phase Serve the Mission?
Classify each phase into one of three categories:
- Aligned — directly advances the north star. The phase would not exist without the mission.
- Tangential — useful work, but not mission-critical. Could be deferred or cut without harming the mission. Examples: documentation cleanup, nice-to-have tooling, speculative features.
- Misaligned — actively pulls away from the mission. Introduces scope, dependencies, or complexity that does not serve the north star. Rare but high-impact when found.
3b: Are Success Metrics Aligned?
If the plan defines validation criteria or success metrics for a phase:
- Do those metrics measure progress toward the mission, or do they measure something else?
- Could a phase pass its own validation criteria while failing to advance the mission?
3c: Is the Phase Proportionate?
- Does the effort size (S/M/L/XL) match the mission importance of the deliverable?
- A large phase for a tangential deliverable is a signal — the plan may be over-investing in non-mission work.
Step 4: Aggregate Drift Check
After assessing individual phases, step back and check the plan as a whole:
- Ratio check: What fraction of the plan's phases are tangential or misaligned? A plan where most phases are aligned but one is tangential is healthy. A plan where half the phases are tangential has drifted.
- Cumulative scope: Do the individually-tangential phases add up to a significant scope commitment? Three small tangential phases may collectively represent more effort than the core mission work.
- Mission coverage: Does the plan address the mission comprehensively, or does it leave mission-critical areas untouched while investing in adjacent concerns?
- Metric alignment: Do the plan's aggregate validation criteria, taken together, confirm the mission will be advanced? Or could all criteria pass while the mission remains unserved?
Step 5: Emit Findings
Classify each finding using the severity vocabulary defined in skills/_shared/contracts/lens-result-contract.md (critical > major > minor > info).
Tag each finding with a concern area. This skill's primary concern area is scope. Secondary: architecture (when misalignment stems from structural choices). Use the closest match from the canonical set in the contract.
Map findings to body sections:
| Section | Typical findings |
|---|---|
| Blockers | Phase actively misaligned with mission; plan's success criteria could pass while mission fails |
| Risks | Aggregate drift — most effort goes to tangential work; mission-critical areas left unaddressed |
| Gaps | Mission coverage holes — areas the mission demands that the plan does not touch |
| Questions | Ambiguous alignment — phase could serve the mission depending on interpretation |
| Observations | Tangential phases that could be deferred; no mission source found (suggest creating one) |
Structured Result Emission (--dispatch only)
After Step 5, emit a single markdown document with YAML frontmatter as the FINAL output. No text before or after this document.
Format: Follow the canonical schema at skills/_shared/contracts/lens-result-contract.md. That contract is the single source of truth for all lens skill --dispatch output.
For this skill, set lens: align-to-mission in frontmatter. All other fields, severity vocabulary, body sections (Blockers/Risks/Gaps/Questions/Observations), concern area values, and blocked/failed output shape are defined in the contract.
Interactive Mode (no --dispatch)
When invoked without --dispatch, this skill is an advisor, not a report generator.
Enter Plan Mode. Call EnterPlanMode. All analysis is read-only.
Execute Steps 1-5 above, then present findings as an advisory conversation:
Advisory Format
Lead with the mission statement (Step 2) and a per-phase alignment summary table. Then present each concern with options and leans.
For each finding, present:
- The concern — what the assessment revealed, stated concretely.
- Options — 2-3 ways the plan could address the concern (including "keep as-is" when the concern is minor).
- Lean — which option you'd pick and why, in one sentence.
- Rationale — the reasoning behind the lean, grounded in the mission.
Example Advisory Output
## Mission Alignment Review: [Plan Title]
### Mission
[The north-star statement, with source noted]
### Phase Alignment Summary
| Phase | Alignment | Notes |
|-------|-----------|-------|
| 2a. [name] | Aligned | Directly serves [mission aspect] |
| 2b. [name] | Tangential | Useful but not mission-critical |
| 2c. [name] | Aligned | ... |
### Aggregate Assessment
[Does the plan as a whole advance the mission? Ratio of aligned/tangential/misaligned.]
### Concerns
**Concern:** Phase 2b is tangential — [explanation]
- **(a)** Defer to a follow-up plan
- **(b)** Keep but reduce scope to [specific cut]
- **(c)** Keep as-is — the tangential work enables future mission-critical work
- **Lean:** (a) — [one-sentence reason grounded in mission]
Use one heading per concern. Phases that are cleanly aligned need only their summary table row — no separate section.
Relationship to /adversarial-review
/adversarial-review touches mission alignment tangentially across its seven lenses but does not dedicate a full pass to it. This standalone skill gives mission alignment a full context window and structured per-phase assessment for independent fleet dispatch by /ironclad.
Red Flags — You Are Doing This Wrong
| Symptom | Problem |
|---|---|
| Every phase is "aligned" | You are pattern-matching on topic similarity, not assessing mission criticality. A phase about the same domain is not automatically mission-aligned. |
| You assessed alignment without stating the mission first | The mission is the anchor. Without it, alignment is opinion. Go back to Step 2. |
| You flagged a phase as misaligned but cannot explain what mission it pulls away from | Misaligned requires a direction — away from what? Restate the conflict concretely. |
| You skipped the aggregate drift check | Individual phases can each look reasonable while the sum drifts. Step 4 is mandatory. |
| You are evaluating implementation quality, not mission alignment | This lens checks what the plan builds, not how well it builds it. Implementation quality is other lenses' job. |
| You blocked on missing PROJECT_MISSION.md | Missing mission source is info-severity, not a blocker. Use the fallback hierarchy or infer from the plan's Goal. |