176 skills · sorted by relevance
Build and evaluate single-agent reinforcement learning experiments, where an agent learns actions from rewards. Use Stable Baselines3 to choose algorithms, create environments, monitor training, and save models with the normalization data needed for consistent evaluation.
Analyze authentication logs to spot suspicious sign-ins, password attacks, and possible account compromise. Statistical baselines and machine learning help compare login activity with normal user behavior, while example SIEM queries support security investigations.
Apply stable-baselines3 in reproducible local data workflows with version-aware APIs, explicit assumptions, and validation. Use when the user chooses stable-baselines3 or its strengths fit the task.
The English summary is being prepared. Showing the original description.
Establishes behavioral baselines from historical telemetry (process, network, or logon events) so hunts can flag rare and first-seen activity instead of relying on static signatures. Activates for requests to build a telemetry baseline, find rare or first-seen activity, or compute frequency baselines for anomaly hunting.
The English summary is being prepared. Showing the original description.
Run a prepared set of test cases to check agent-booster success rates and routing decisions. Optionally compare results with Gemini or Claude models, then save detailed measurements and summaries as JSON records for later verification.
Audit Azure Storage settings for public containers, encryption, access permissions, network restrictions, and logging gaps. Use command-line checks and monitoring alerts to produce findings and remediation guidance across subscriptions.
Configures and enforces SwiftLint in Swift projects using build tool plugins, run scripts, and CI. Covers .swiftlint.yml configuration, disabled_rules, opt_in_rules, only_rules, analyzer_rules, baselines, autocorrect, swiftlint:disable suppressions, reporter formats (sarif, json, checkstyle), strict and lenient modes, SwiftLintBuildToolPlugin via SimplyDanny/SwiftLintPlugins, swift package plugin swiftlint, Xcode run script phases, CI integration, multiple configuration files, and rollout strategies for existing codebases. Use when setting up SwiftLint, configuring lint rules, suppressing warnings, creating baselines, choosing between build tool plugin and run script, or integrating SwiftLint into CI.
The English summary is being prepared. Showing the original description.
Build the deterministic evidence ledger (artifact_manifest.json + claims.json) that every other Anti-Autoresearch auditor reads. One pass inventories artifacts, derives the observability level (L0 PDF-only / L1 +LaTeX / L2 +repo+results) by fixed rule, and extracts span-anchored, hashed, checkable claims (numbers, comparisons, scope, method, baselines, citations, captions, table cells) into claims.json. An OPTIONAL additive cross-model pass ADDS span-anchored semantic claims — method, theorem statements with their assumptions, definitions, proof/derivation steps and equations, scope, baselines, conclusions, the motivation span, and reproducibility-artifact references (the proof, derivation, and structure anchors the family B/D/G auditors need) — it never invents a number, emits a finding, or computes a verdict. Run FIRST, before any audit skill. Triggers: "build the ledger", "extract claims", "prep for integrity audit", "evidence ledger", "建证据账本".
The English summary is being prepared. Showing the original description.
Audit existing performance testing — find missing benchmarks, stale baselines, and CI gaps. Use when asked "do we have benchmarks", "audit our performance testing", or "are our baselines stale".
The English summary is being prepared. Showing the original description.
Establish and maintain longitudinal baselines of CLI binary contents across versions. Covers marker selection by category (API / identity / config / telemetry / flag / function), weighted scoring, threshold-based system-presence detection, and per-version baseline records. Use when tracking a feature's lifecycle across releases, when probing for dark-launched or removed capabilities, or when verifying that a scanning tool itself still catches known-good markers on old binaries.
The English summary is being prepared. Showing the original description.
Knowledge base from NIST SP 800-37 Revision 2, the Risk Management Framework (RMF) for information systems and organizations. Use for the seven RMF steps (Prepare, Categorize, Select, Implement, Assess, Authorize, Monitor) and their tasks; the three-tier organization-wide risk model; authorization boundaries; the requirements-to-controls relationship; security categorization (FIPS 199/200 high-water mark) and control selection/tailoring (SP 800-53/53B baselines); assessment, the authorization decision and risk acceptance, continuous monitoring and ongoing authorization; the security/privacy distinction; and supply chain risk management. This is the security-and-privacy-risk-governance edge of systems engineering — it governs how much protection and whether to accept residual risk, complementing the disciplines that decide how protection is built. Does NOT reproduce the SP 800-53 control catalog or its baselines, the FIPS impact tables, or the ISO/IEC/IEEE 15288 process text; it names and routes to them rather than restating them.
The English summary is being prepared. Showing the original description.
Measure web page speed, API response times, and build or test duration. Save baseline measurements as JSON tracked in Git, then compare results before and after changes to identify performance regressions and assess launch targets.
Measure web page loading times, rendering milestones, and resource sizes in a real browser. Save baseline measurements, compare results before and after code changes, and inspect historical trends to spot performance regressions.
Monitor a live app after deployment for console errors, slower performance, and page failures. Periodic screenshots and comparisons with pre-deployment baselines help identify anomalies and check whether the release is behaving as expected.
Summarizes recorded system metrics for a chosen period, including completed tasks, errors, memory use, and token consumption. It compares values with historical baselines and flags unusual changes to help assess agent team health.
Benchmark security detection systems against 10 known vulnerabilities and nine harmless decoys. Compare four approaches, including an evolved champion, and receive structured JSON covering detection, false alarms, patch tests, reproduction, unsafe outputs, and overall pass or fail.
Compare actual AI session costs with estimates for using only Haiku, Sonnet, or Opus. The skill reports savings amounts and percentages, helping teams assess model selection costs while leaving differences in answer quality and retries outside the calculation.
Evaluate 3D point clouds and movement trajectories against reference data with CloudAnalyzer. Prepare inputs, measure differences, apply quality thresholds, inspect results visually, and track baseline history for mapping, localization, and perception workflows.
Prepare an evaluation framework before fine-tuning a language model. Build labeled examples, checks for each failure type, and a base-model benchmark, while calibrating model-based judges against human labels to support comparisons after training.
Evaluate LLM applications using automated scores, human reviews, and model-based judging. Build repeatable checks to compare models or prompts, assess answer quality, and detect declines in performance before deployment.
Customize AgentRC readiness assessments to match team or organization standards. Choose or create policies that adjust check importance, disable irrelevant checks, set pass thresholds, and define requirements for automated checks in CI workflows.
Monitor Modbus traffic in industrial control networks to identify unauthorized writes, unusual commands, and departures from normal communication. Use packet analysis and intrusion detection rules to investigate suspected controller changes and produce reports with recommended actions.
Analyze endpoint, cloud storage, and email logs for signs of insider data theft. Use pandas to compare upload volumes with normal user behavior and examine data loss prevention alerts, unusual file access, and activity outside working hours.
Build or tune a monitoring pipeline that learns normal user behavior from activity logs. Use Elasticsearch or OpenSearch to score unusual activity, compare colleagues in similar roles, and combine warning signs into insider threat alerts.