Nicholls Inc
Official@nicholls-inc · Canada
Offers formal verification, software assurance, and rigorous specification modeling for high-integrity codebases using Lean 4 and Dafny.
Agent Skills by Nicholls Inc
Showing 29 vetted skills indexed across 1 GitHub repositories.
suggest-prompts
Fetch curated GitHub Copilot prompts and compare local versions for updates.
suggest-instructions
Fetch and compare GitHub Copilot instruction files against local repository files.
suggest-skills
Analyze repository content to recommend relevant GitHub Copilot agent skills.
assurance-roadmap-check
Validate assurance roadmap item statuses against repository activity and documentation.
lean-spec
Convert signed-off informal verification specs into type-checked Lean 4 stubs.
acceptance-oracle-draft
Generate YAML/JSON acceptance scenarios with CI integration for automated QA.
extract-code
Convert verified Dafny code into Python or Go.
intent-check
Verify code diffs against documented invariants to prevent regressions.
lean-impl
Translate Python, Go, or Dafny source code into formal Lean 4 models.
correspondence-review
Classify Lean definitions against source code as exact, abstraction, approximation, or mismatch.
lightweight-verify
Generate assertions, invariants, and property-based tests for functions.
drt-oracle
Detect inconsistencies between Lean formal models and production implementations via randomized testing.
compare-patches
Analyze two code patches for semantic equivalence using structure and tests.
suggest-specs
Identify code functions and modules that benefit from formal specification generation.
trace-execution
Trace code execution paths and build call graphs from entry points.
spec-adversary
Propose undocumented invariants for module specifications in software assurance workflows.
assurance-probe
Run mutation-based evaluation of Python property invariants and report weak tests via GitHub issues.
assurance-layer-audit
Audit a code repository's tooling and verification signals against the formal assurance hierarchy.
protected-surface-amend
Generate a governance amendment record for protected surface changes.
assurance-init
Create assurance governance files and directories in a code repository.
locate-fault
Trace code paths and predict root causes of failing tests.
assurance-status
Verify repository assurance documentation presence and analyze drift and coverage issues.
generate-verified
Generate Dafny implementations that satisfy specified formal properties.
check-regressions
Detect Dafny specification changes requiring re-verification via git diffs and source hashes.
Frequently Asked Questions About Nicholls Inc
FAQPage SchemaWhat specific tasks can I perform using Nicholls Inc skills?▼
You can perform formal verification of source code, generate property-based tests, translate natural language requirements into Dafny specifications, and audit repository governance documentation. These capabilities enable rigorous semantic analysis and ensure code correctness against documented invariants.
Which personas benefit most from these verification capabilities?▼
Software engineers, security researchers, and compliance officers working on high-integrity or safety-critical systems benefit most. These skills are designed for developers needing to bridge the gap between informal requirements and mathematically verified implementations.
What are the prerequisites for implementing these verification models?▼
Implementation requires a repository containing source code in Python, Go, or Dafny, alongside established formal verification environments like Lean 4. Users must define initial invariants or natural language specifications to enable the generation of formal models and property-based tests.