Skill UI

Builds phase-based legal fee budgets, monitors WIP against budget, explains variance with root cause analysis, forecasts to complete, manages alternative fee arrangements (fixed/capped fees), models scope change financial impact, and supports realisation and write-off analysis for legal matters.

Diagnose and fix failures in built .ppmplugin controls: crashes, silent no-ops, PCF errors, or incorrect behavior. Investigates symptoms through error codes, manifests, native modules, and dispatch to locate root causes, then applies approved surgical fixes. Use when a control breaks on device.

Debug a failed Power Automate flow run. Use when a flow failed, has errors, or the user wants to troubleshoot a run. Identifies root cause, classifies errors, and suggests fixes.

Wraps DeepSeek's deepseek-reasoner model to expose chain-of-thought reasoning separately from the final answer. Enables multi-step reasoning tasks such as proofs, planning, root-cause analysis, and bug triage with explicit thinking output for audit or verification. Handles API key via environment, degrades gracefully when unavailable.

Analyze OFAC or OFSI enforcement actions from URLs, pasted text, or uploaded documents and generate a structured root cause analysis Excel (.xlsx) file. The output table covers root causes, what went wrong, how it went wrong, preventive controls, and organizational immunity checks for compliance teams, counsel, and consultants.

A structured debugging skill for the googleapis/mcp-toolbox repository. It guides reading the actual error, establishing what changed, reproducing the failure, isolating the root cause, and then fixing that cause rather than the symptom. Use it when tests or CI jobs are red, a build breaks, many packages fail, or a test passes locally but fails in CI.

Audit defect backlogs and resolve issue clusters by verified root cause: build read-only mechanism maps, rank deletion-first fixes, and close issues only with rule-backed evidence.

A skill for triaging issues by root cause, avoiding symptom-by-symptom fixes, and designing solutions that simplify systems. It provides hard rules for classification, test evidence, and over-engineering prevention, suitable for maintainers handling bug reports and backlogs.

Close defects and incidents with a structured BUG RECEIPT output including VERIFIED, PARTIAL, or BLOCKED status. Establishes evidence boundary, traces root cause, makes responsible changes, and closes the proof loop with explicit checks and gaps.

Forensic analyzer for AI coding sessions. Identifies root causes of scope drift and rework by classifying user specs, agent errors, and repo friction. Generates evidence-backed reports to improve future prompts and project health.

This skill prevents AI coding agents from entering ineffective repair loops by establishing a bounded repair contract. It uses stable failure fingerprints, a three-attempt budget, and real-path verification to ensure progress. When fixes do not change observable states, it mandates shifting root-cause strategies rather than retrying. Includes rollback controls and negative testing to validate actual system behavior over agent claims.

An evidence-first procedure for system performance debugging, root-cause analysis and incident reports. Applies Brendan Gregg's USE and TSA methods to sweep resources and thread states, then drills down with off-CPU analysis and flame graphs, ending in triage notes, RCA reports or blameless postmortems.