Use when starting any grafana-app-sdk work — scaffolding a Grafana app, initializing a Grafana App Platform app, picking a deployment mode (standalone operator / grafana/apps / frontend-only), wiring app-specific config, or onboarding to the SDK. Covers `grafana-app-sdk` CLI install, `project init` per deployment mode, project layout, the schema-centric workflow (CUE kinds → generated code → reconciler/admission logic), and `SpecificConfig` for env-driven app configuration. Use even if the user just says "make a new app", "Grafana app platform", "kinds and watchers", "operator scaffold", or asks about `apps/` inside the Grafana repo, without naming the SDK explicitly.

Analyze the impact of upgrading a dependency before you do it. Use when the user says "should I upgrade lodash", "what breaks if I update express", "upgrade impact", "endor upgrade", "breaking changes from upgrading", or wants to find the safest version that fixes vulnerabilities. Uses pre-computed Endor Labs data — no scanning required. Do NOT use for just checking vulnerabilities (/endor-check) or applying a fix (/endor-fix).

Display and filter security findings from Endor Labs. Use when the user says "show findings", "list vulnerabilities", "what did the scan find", "endor findings", "show me critical reachable vulns", or wants to browse/filter results after a scan. Supports filtering by severity, reachability, category (vuln/sast/secrets/license). Do NOT use for running a new scan (/endor-scan) or explaining a specific CVE (/endor-explain).

Check if a specific dependency has known vulnerabilities or malware using Endor Labs. Use when the user names a package and wants to know if it's safe, says "check lodash", "is express vulnerable", "any CVEs in django", "endor check", "is this package safe", or provides a package name after installing a dependency. Do NOT use for scanning an entire repo (/endor-scan) or viewing existing findings (/endor-findings).

Core building blocks of Suricata signatures and multi-condition DPI logic

Automatically generate regression tests for Java codebases by analyzing changes between old and new code versions. Use when users need to: (1) Generate tests after refactoring or code changes, (2) Ensure previously tested behavior still works in new versions, (3) Cover modified or newly added code paths, (4) Migrate existing tests to work with updated APIs or signatures, (5) Maintain test coverage during code evolution. Supports JUnit and TestNG frameworks with unit tests, parameterized tests, and exception testing patterns.

Algolia launch, regression, and implementation QA. Use when auditing an Algolia build before release, validating relevance changes, checking search UI behavior, verifying event instrumentation, diagnosing analytics gaps, reviewing secured keys, or creating a smoke test plan for search, browse, autocomplete, indexing, and configuration changes. Do NOT use to make live account changes or inspect live account data; use algolia-mcp or algolia-cli first. Do NOT use as the primary implementation skill when the task is clearly data modeling, events setup, InstantSearch, Autocomplete, NeuralSearch, or Agent Studio; use the focused skill first and this skill for final QA.

Work with Obsidian vaults (plain Markdown notes) and automate via obsidian-cli.

Audit a workspace for recurring friction, pick the highest-leverage one, and build the automation that removes it. Use when the user asks "what should I automate", "what's wasting my time", "find the friction in my workflow", wants a self-improvement pass over their own tooling, or runs this on a recurring loop.

Implements a use case's backend: FastAPI router, Pydantic request/response schemas, service/domain layer, and SQLAlchemy models/queries. Use when the user asks to "implement the backend for UC-xxx", "build the API endpoint", "add a FastAPI route", "write the service layer", or mentions FastAPI implementation, Pydantic schemas, or SQLAlchemy queries for a specific use case.

Evaluate a repo agent harness (AGENTS.md, rules, skills, skill refs) for broken paths/commands, redundant instructions, and usefulness using a stack-agnostic dual-judge protocol with planted traps. HIGH PRIORITY questionnaires at top: Q1 optional docs, Q2 B/C budget before Track A (certainty/tokens). A always runs after Q2; B/C opt-in. ADRs/RFCs excluded from T2. Mixed apply uses 11-mixed-apply.md (KEEP/CUT). Use when the user says harness eval, harness-eval, harness debug, audit AGENTS.md, audit skills/rules, instruction audit, redundancy of agent instructions, usefulness of skills, Ship/Review/Hold/Slim/Keep-core for harness, or wants Track A/B/C harness evaluation. Do NOT use for harness setup or init, feature spec-driven work (tlc-spec-driven), or applying Ship/Slim trims unless the user explicitly asks after the report.

Translates natural-language or pseudocode descriptions of concurrent and distributed systems into TLA+ specifications ready for the TLC model checker. Identifies state variables, actions, type invariants, safety properties, and liveness properties from the description. Use when formalizing a protocol, when the user describes a distributed algorithm to verify, when designing a consensus or locking scheme, or when starting formal verification of a concurrent system.

Translates Python functions into Dafny, adding types, pre/postconditions, and loop invariants sufficient for Dafny to verify. Use when formally verifying a Python algorithm, when the user wants machine-checked correctness for a critical function, or when building a verified reference implementation.

Extracts a TLA+ specification from concurrent or distributed code, modeling the state machine, actions, and fairness conditions for model checking with TLC. Use when verifying concurrency properties of production code, when designing a protocol and wanting to check it before implementation, or when the user has a race condition and needs to prove the fix.

Infers likely loop invariants and function contracts by observing execution traces, synthesizing candidates, and checking them inductively. Use when a verifier rejects a loop because the invariant is missing or too weak, when a Daikon-style tool is needed, or before translating code to a verification language.

Generates unit tests for a function or class by analyzing branches, boundaries, and error paths — then emits test code in the project's existing framework and style. Covers happy path, edge cases, and failure modes with mocks for external dependencies. Use when writing tests for new code, when backfilling coverage on untested functions, when the user asks to generate tests, or when a coverage report shows specific gaps.

Finds and removes redundant tests — tests that cover the same code, kill the same mutants, or assert the same behavior — to shrink suite runtime without losing coverage. Use when the test suite is slow, when tests have accumulated over years of copy-paste, or when CI costs are too high.

Updates broken pytest tests after intentional code changes — triaging assertion failures from mock-coupling failures from genuine regressions, using Python's introspection to automate where safe. Use when a refactor or API change leaves a pile of failing tests and you need to decide update vs. fix vs. delete.

Uses mutation testing to find weak assertions and missing tests — injects small bugs and checks if the suite catches them, then generates tests targeting the surviving mutants. Use when coverage is high but bugs still ship, when auditing test quality, or when deciding if the suite is good enough.

Updates broken JUnit tests after a deliberate code change — distinguishing tests that broke because the behavior changed (update assertion) from tests that broke because they were overcoupled to structure (loosen or delete). Use after API changes, refactors, or intentional behavior changes leave a trail of failing tests.