agentleFS
nasanasa/fprime/.github/skills/fprime-component-checklist-review/SKILL.md

fprime-component-checklist-review

Skill12k starsChanged 21 months agoApache-2.0View on GitHub

Fill in the F Prime Component Development Checklist (v1.2.0) for a component from automated evidence plus a short LLM review of the items no tool can prove. Maps every checklist item to its CodeQL query, component-checks script, review rule, or human step; defines the review rules CL-1..CL-8 for the residual items. Use when asked to run, fill, or audit the component checklist, or to review a component for flight readiness. Keywords: F Prime, checklist, component, flight software, CodeQL, component-checks, close-out, ISR, units, boundary tests.

Trending

More talk lately.about the whole repository

Dec 2025: 0 mentionsJan 2026: 0 mentionsFeb 2026: 0 mentionsMar 2026: 1 mentionApr 2026: 0 mentionsMay 2026: 22 mentionsJun 2026: 3 mentionsJul 2026: 7 mentionsAug 2026: 2 mentionsSep 2026: 7 mentionsOct 2026: 11 mentionsNov 2026: 1 mention
Dec 2025Jun 2026Nov 2026
Mentions a month of the nasa/fprime repository that our weekly crawl found. There is no adoption data for this file yet.

Who's talking about nasa/fprime

The biggest 5 of 71 mentions linking to the repository that our crawl found. Each line is their own headline.

See all 71

Talked about on GitHub, YouTube, Hacker News and the web (71 times).

On GitHub · 50

45 more on GitHub

On YouTube · 13

8 more on YouTube

On the web · 2

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 1 is distinctive, and 2 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    N/A: only when the mapped check reports SKIP with a reason or the component has none of the construct (e.g.

    line 31
  2. Distinctive:

    CL-3 Units in names (3.33): Every variable, member, parameter, FPP telemetry channel, and command argument that carries a physical quantity …

    line 148
  3. Very specific:

    Run component-checks for the module (or read its published checks/summary.json) and the CodeQL SARIF for the branch.

    line 176

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

F Prime component checklist review

The checklist has 73 items in five sections. All but one are decided by tools; this skill tells you where each verdict comes from and defines the few judgements that remain for the reviewer.

Evidence sources

SourceWhereProduces
CodeQL (fprime/jpl-c-queries).github/codeql/fprime-queries/*.ql, run by codeql-jpl-standard.ymlSARIF alerts per rule id cpp/fprime/...
component-checksnasa/fprime-actions/component-checks, run by component-checks.yml<module>/checks/summary.json with ids R1..R9, D1..D10, I1..I4, U1..U5, C1..C3
This skillrules CL-1..CL-8 belowreviewer findings

A checklist item is YES when every mapped rule/check passes for the module, NO when any fails (cite the alert or check line in Notes), N/A only when the mapped check reports SKIP with a reason or the component has none of the construct (e.g. no ISR, no floats).

Item map

Legend: Q = CodeQL rule id (cpp/fprime/ prefix omitted), S = component-checks id, L = review rule in this skill, H = human.

1 Requirements

ItemMechanism
1.1 behavioursS R9 (Kind column covers Behavior) + R1 (commands) + L CL-7
1.2 inputsS R9 (Input) + R6 (ports) + R4 (parameters)
1.3 outputsS R9 (Output) + R2 (telemetry) + R3 (events) + R5 (data products)
1.4 memoryS R9 (Memory, or waived in SDD)
1.5 hardware I/OS R9 (Hardware I/O, or waived)
1.6 performanceS R9 (Performance, or waived)
1.7 verification methodS R8
1.8 external ICD/FDDS D9 (external-interfaces section or explicit None)

2 Design

ItemMechanism
2.1 high-level designS check_sdd.py grade >= silver + L CL-8
2.2 block diagramS D9
2.3 port table vs modelS D7
2.4 input-port behaviourS D10 + L CL-8
2.5 port handlersS D1
2.6 command handlersS D2
2.7 sequence diagramsS D9
2.8 state machines / use casesS D9
2.9 state-machine actionsS D3
2.10 cmd/tlm/evr/prm tablesS D8
2.11 data-product tablesS D8
2.12 DP default priorityS D4
2.13 every port has a use caseS R6 + D10
2.14 port directions/typesS D7
2.15 component kindS D9
2.16 requirements mapped to designS D5
2.17 FPP compilesS D6

3 Implementation

ItemMechanism
3.1 CMake dependenciesS I1
3.2 no warningsS I2
3.3 loops boundedQ jpl-c/loop-bounds
3.4 cyclic messages gatedQ unthrottled-cyclic-event
3.5 switch default/breakQ switch-default, switch-case-fallthrough
3.6 fallthrough annotatedQ switch-case-fallthrough
3.7 variables localQ global-mutable-variable
3.8 access specifiersQ public-data-member
3.9 float equalityQ float-equality
3.10 float precision / margins / loopsQ float-loop-control, float-narrowing + L CL-6
3.11 initialised before useQ uninitialized-member + stock cpp/uninitialized-local
3.12 DP priority configurableS I3
3.13 command acknowledgedQ command-response
3.14 validation failure returnsQ validation-failure-return
3.15 internal port invoke()Q internal-port-invoke
3.16 pointer args assertedQ checking-parameter-values
3.17 state assertedQ jpl-c/use-of-assertions-density
3.18 no assert on HW/user inputQ assert-on-untrusted-input
3.19 ISR no floatQ no-float-in-isr
3.20 ISR minimalQ isr-complexity
3.21 ISR state before enableL CL-1
3.22 ISR non-blockingQ no-blocking-in-isr
3.23 interrupt enables cleared at initL CL-2
3.24 rate-group handlers non-blockingQ no-blocking-in-sched-handler
3.25 no file I/O in critical codeQ no-file-access-in-critical
3.26 compound statementsQ compound-statements
3.27 division guardedQ division-guard, float-division-guard
3.28 array boundsQ unchecked-array-index, checking-parameter-values
3.29 no nested function declQ no-function-decl-in-function
3.30 for-header side effectsQ for-header-side-effects
3.31 signed overflowQ signed-overflow
3.32 return values checkedQ jpl-c/checking-return-values
3.33 units in namesL CL-3
3.34 named constantsQ magic-numbers, define-constant
3.35 no printfQ no-printf
3.36 fixed-width typesQ jpl-c/basic-int-types
3.37 static analysisS I4
3.38 implementation matches reqs/designL CL-4

4 Unit testing

ItemMechanism
4.1 tests publishedS U1
4.2 UT CMake dependenciesS U3
4.3 boundary testsL CL-5
4.4 no leaksS U4
4.5 coverageS U5

5 Close-out

ItemMechanism
5.1 no TBD/TODO/printfQ incomplete-code, no-printf
5.2 topology buildsS C1
5.3 runs on hardwareH performs the run; S C2 verifies docs/hardware-run.md
5.4 SDD finalS D7, D8, D9, check_sdd.py
5.5 CI passesS C3

Review rules (residual LLM items)

Apply each rule only to the construct it names; report N/A when the component has none. Cite file:line for every finding.

  • CL-1 ISR state before enable (3.21). For every function matched by the ISR naming convention (*_isr, *interrupt_handler*, *_irq*) find the code that enables its interrupt (register write, request_irq, enableInterrupt, vector install). Every variable the ISR reads or writes must be assigned on a path that completes before that enable call. Finding: ISR state uninitialised (severity error).
  • CL-2 Interrupt enables cleared at init (3.23). In the component's constructor/init/configure/preamble, interrupt-enable bits must be explicitly cleared (or interrupts masked) before peripheral configuration, not assumed clear from reset. Finding: Interrupt enable assumed clear (warning).
  • CL-3 Units in names (3.33). Every variable, member, parameter, FPP telemetry channel, and command argument that carries a physical quantity (time, distance, angle, rate, temperature, voltage, current, size in bytes vs words) must encode the unit in its name (timeoutMs, angleRad, sizeBytes) or in an FPP @ annotation. Finding: Missing unit in name (recommendation).
  • CL-4 Implementation matches requirements and design (3.38). For each requirement row, locate the handler(s) that satisfy it and confirm the behaviour, including each error path the SDD describes. Report any requirement with no implementing code, and any handler behaviour the SDD does not describe. Finding: Implementation diverges from SDD (error).
  • CL-5 Boundary tests (4.3). For each bounded quantity in the model (array sizes, FW_ASSERT limits, configuration constants, command-argument ranges, queue depths) there must be a unit test exercising the limit value and one past it. Finding: Boundary untested (warning).
  • CL-6 Float precision justified (3.10). Each F32 quantity must have sufficient precision for its range (e.g. seconds since epoch in F32 loses seconds after ~190 days). Flag accumulators and time values held in F32. Finding: Insufficient float precision (warning).
  • CL-7 Requirement quality (1.1). Apply fprime-requirements-review to the requirements table when it has changed in the PR.
  • CL-8 Design prose adequacy (2.1, 2.4). The SDD must describe, for each input port and command, what the component does, including the rejection path. Table rows alone do not satisfy this. Finding: Behaviour not described (warning).

Procedure

  1. Run component-checks for the module (or read its published checks/summary.json) and the CodeQL SARIF for the branch.
  2. Fill sections 1, 2, 4, 5 and the Q/S rows of section 3 from that evidence.
  3. Apply CL-1..CL-8; fill the remaining rows.
  4. Item 5.3: confirm docs/hardware-run.md (checked by C2) names a human tester; the run itself is not delegable.
  5. Emit the checklist with status (YES/NO/N/A) and Notes per the checklist's "Status Legend"; non-YES rows must carry the alert or check line that justifies them.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.