CC fda-med-device
Reconstructed reference notes on US FDA medical-device software regulation, built from seven public-domain FDA sources (the QMSR final rule 89 FR 7496, FDA device software guidance documents, and the eSTAR program page; pinned 2026-10-12). Use for regulatory pathways and premarket submissions (510(k), De Novo, PMA, eSTAR), premarket software documentation levels (Basic or Enhanced), software validation and Computer Software Assurance, the Quality Management System Regulation (QMSR, 21 CFR 820), medical-device cybersecurity (secure product development, SBOMs, FD&C Act section 524B, labelling), clinical decision support boundaries, off-the-shelf software, MDDS and other software-function scope questions, and the 510(k) software-change trigger. SCOPE LIMITS: US FDA device regulation only; no EU MDR or UKCA content (see `med-device-eu`); no ISO 13485, ISO 14971, or IEC 62304 text, which are named and summarized only as FDA obligations; no drug, biologic, or combination-product rules; reflects the pinned source dates, not later FDA actions; synthesized reference notes, not legal advice and not a substitute for the guidance documents. LICENCE: Public Domain (US Government work, 17 U.S.C. 105).
Reconstructed reference notes on US FDA medical-device software regulation, built from seven public-domain FDA sources (the QMSR final rule 89 FR 7496, FDA…
As a process C 64/100 · Has gaps — weak spots: result and completion, failures and branches, running it twice
How to improve
- Shorten the description to 1024 characters.
- Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
- A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.
Guard findings · 0
✓ No critical or high findings
Files scanned: 17. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- error
description-longdescription is 1206 chars, limit 1024 - note
edit-residuethe text marks something as outdated (lines 71): check that old rules are not kept next to new ones — the full check reads the text for contradictions
Process rating: all ten parameters 64/100
- 0Result and completion. Does not say what the result is
- 0Failures and branches. Linear process with no failure handling
- 30Running it twice. 1 mutating operations with no state check
- 50When it triggers. No condition that starts the skill
- 70Inputs and preconditions. Inputs and preconditions are listed
- 100Tools and files. No external tools needed
- 100Steps. 39 steps
- 100Consistency. Name and required fields are in place
- 100Execution cost. Instruction body is 3604 tokens
- 100Progress reporting. Reports progress
Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.
Quality signals
- +5Description has no quoted example phrases that should trigger the skill
- +4Description does not say when NOT to use the skill (false activations)
- +3Description length 1206: 120–800 characters recommended
- +3Output format is not stated: the model decides each time
- +4No input/output examples
- +2Single-language instructions
- +4Structure: 16 headings
- +3Step-by-step instructions: 39 items
- +1License stated
Quality base 70; lint remarks subtract, signals add up to 100. Result: 53.