Summary
Create a procedure (skill) for assessing the current state of the endeavor and producing a prioritized recommendation for what to do next, with reasoning grounded in goals, satisfaction deficit, and leverage. This is the “take a better look at where we are” capability that the method currently lacks.
Motivation
The review-plans skill (v0.3) does mechanical sorting: group by status, sort by priority, check dependencies, warn on WIP violations. It produces a board view and a flat recommendation. But when three proposed plans have the same priority and no blocking dependencies, it cannot distinguish between them. It reports the landscape; it does not assess it.
The gap surfaced concretely: after completing plan 0017, three proposed plans (0022, 0023, 0024) were pitched to emsenn with no strong basis for choosing one over the others. The murkiness in that choice is a failure in the method — the procedure for “determine what to do next” stops at mechanical sorting and does not continue into evaluative assessment.
A situational assessment procedure would:
- Evaluate progress toward each goal (not just plan counts, but whether recent completions moved the needle)
- Identify the method-practice gap: where is the method not being followed? where is practice outrunning method?
- Apply leverage analysis (policy 009): which action, if completed, makes the most other actions easier or unnecessary?
- Produce a recommendation with reasoning, not just a sorted list
This is the difference between a dashboard and an assessment. Dashboards show data; assessments interpret data against objectives.
Steps
- Write a situational-assessment skill at
plans/skills/situational-assessment/SKILL.md. - The skill reads: all goals (horizon, key results, completion state), all active/proposed/accepted plans, recent plan completions, the method-practice gap (what’s specified but not practiced, what’s practiced but not specified).
- The skill produces: a state-of-the-endeavor summary (2-3 paragraphs), a leverage-ranked list of candidate next actions with reasoning for each, a recommended next action with justification, and identified method-practice gaps.
- Distinguish from review-plans: review-plans is mechanical (sort, group, warn). Situational assessment is evaluative (interpret, compare to goals, recommend with reasoning). Review-plans feeds data into the assessment; the assessment is the interpretation.
- Test by running the assessment against the current state and checking whether the recommendation would have resolved the murkiness that prompted this plan.
Done when
- Situational assessment skill exists with clear instructions
- The skill produces a recommendation with reasoning (not just a sorted list)
- Running it against current state produces a non-obvious insight (i.e., the recommendation differs from what mechanical sorting would produce)
Dependencies
None. (review-plans provides input data, but it already exists.)
Log
2026-03-08 — Created. Prompted by the inability to choose between three proposed plans (0022, 0023, 0024) after completing plan 0017. The murkiness in that choice exposed a gap: no procedure for evaluative assessment of the endeavor’s state. Policy 007 (build then fill gaps) applies: the task (choose next plan) revealed a gap (no assessment procedure), so we fill the gap first.
2026-03-08 — Activated. Wrote situational-assessment skill v0.1 at plans/skills/situational-assessment/SKILL.md. Key design: five-factor leverage scoring (unblocks, serves near goal, closes gap, improves decisions, ready now). Distinguishes from review-plans: review-plans is mechanical (sort, group, warn); this skill is evaluative (interpret against goals, identify method-practice gaps, rank by leverage).
2026-03-08 — Tested against current state. Non-obvious finding: all 5 goals are still draft — no strategic direction has been formally accepted by emsenn. This is the largest method-practice gap (the lifecycle spec requires emsenn’s goal review, and it hasn’t happened). The leverage-ranked recommendation — review goals before more plan-level work — differs from what mechanical sorting would produce.
2026-03-08 — Completed. All acceptance criteria met:
- Skill exists with clear instructions (7 sections, decision protocols, output formats, tie-breaking procedure)
- Produces recommendation with reasoning (leverage ranking + why-not-alternatives + what-changes-after)
- Test run produced non-obvious insight (goal acceptance gap is higher leverage than any individual proposed plan)