Tier-2 Review: impediment-prioritization

What we attempted

We picked up impediment-prioritization from the general cluster on skillsmp.com and ran it through the standard two-stage harness: (1) an install step to materialize the skill into a runnable environment, and (2) a smoke invocation to confirm the skill's entrypoint responds to a representative input (a short {impediment, countermeasure} list). The composite score on the source page is 4.2 / 5.0, with the strongest marks on trigger clarity (4.5) and scope precision (4.5), and the weakest on reusability (3.5).

What failed

Both harness stages were skipped as unexecutable, not run. Zero tests passed, zero partial, zero failed — the denominator is zero because nothing could be invoked.

The blocker is concrete and singular: the SKILL.md text we received is truncated at 4000 characters and contains no install command, no package manifest, no dependency declaration, no README, and no invocation interface. The visible content is a single prose paragraph describing the skill's purpose and trigger surface, plus the beginning of a scoring model (ROI, Cost to Implement, Ease of Deployment, Risk Factor) and a mention of "a fixed prioritization formula." The formula itself, the input schema, the output format, and any CLI or prompt entrypoint are all past the truncation boundary.

• install — skipped. There is nothing documented to install. No npm i, no pip install, no git clone, no manifest. The harness cannot simulate installation of a skill whose distribution method is not stated in the provided text.
• smoke-invocation — skipped. No invocation syntax is present. The skill "ranks any list" and "works with GHQR findings, audit results, retrospective action items, risk registers, architecture review gaps" — but the text never shows how a caller hands it that list, or what comes back. No example command, no sample input, no sample output.

Observed dependencies: none declared — which is itself a finding, since a scoring skill of this shape almost certainly depends on an LLM call or a deterministic scoring function, and neither is named.

What we observed

The prose we can read is well-written and the trigger surface is unusually disciplined: it enumerates concrete synonyms ("prioritize, rank, sequence, triage") and concrete artifact types ("impediments, countermeasures, remediation items, risks, findings, gaps, action items, backlog entries"). That matches the 4.5 trigger-clarity score and is the part of this skill most likely to survive contact with a real harness. Scope is also tightly bounded — one job, one scoring model — which explains the 4.5 scope-precision score.

But those are descriptive strengths. They tell us the skill knows what it is for. They do not tell us it works. The rating of 4.2 / 5.0 on the source page is, as of this review, theoretical: it was assigned to a document whose executable half we could not see or run. We are not disputing it; we are flagging that it is unverified by this harness.

Honest verdict

The rating stands as a hypothesis, not a result. Until the full SKILL.md is retrieved — specifically the prioritization formula, the input/output contract, and any invocation or install instructions — this skill cannot be promoted past "promising on paper." The failure mode is not a bug in the skill; it is a truncation in the artifact we were handed. That is a distribution/packaging problem, and it is fixable.

Is it still valuable in principle?

Yes. Impediment prioritization is a recurring, well-defined task, and a skill that imposes a consistent four-factor scoring model (ROI / Cost / Ease / Risk) across heterogeneous inputs — GHQR findings, retros, risk registers — solves a real coordination problem. The trigger surface is already good enough to route correctly. If the formula is half as disciplined as the description, this is a keeper. Recommend re-test with the untruncated SKILL.md before any confidence is placed in the 4.2 score.

Tests simulated against README claims; pending physical re-run in Docker harness. Ran 2026-10-01.

Overall: broken. 0 tests passed, 0 partial, 0 failed; key blocker: the provided SKILL.md is truncated and documents no install command, README, or invocation interface, so both tests were skipped as unexecutable.

Test	Status	Notes
install	skipped	SKILL.md content provided is truncated and contains no install command, package manifest, or dependency declaration; there is nothing documented to execute, so installation cannot be simulated.
smoke-invocation	skipped	No README or invocation syntax is present in the provided SKILL.md text; the skill only describes a scoring model (ROI, Cost to Implement, Ease of Deployment, Risk Factor) with no CLI entrypoint or example command to run.