ai

ALWAYS use this skill whenever you would otherwise ask the user a question in free-form chat -- for clarifications, confirmations (especially destructive actions), missing parameters, multiple-choice decisions, or structured form input. Elicitations are routed through the Unique AI Platform UI via `unique-cli elicit create` + `elicit wait` (or `elicit ask` outside an agent harness) so the user gets a proper structured prompt and you get a structured answer back. Do NOT ask the user in plain chat when you can use this skill instead.

Read Agentic Table (magic table / due-diligence) sheets through the unique-cli agentic-table command. Use when the user or task involves inspecting an Agentic Table: a sheet's state and metadata, a specific cell's value or lock state, a cell's edit/answer history, or the sheet's generated export artifacts (reports, question exports). These are read-only (Tier 0) commands — they never modify the sheet and never require confirmation. Access to each sheet is enforced by the platform; a denial is reported as `agentic-table: permission denied`.

Call MCP (Model Context Protocol) server tools on the Unique AI Platform using the unique-cli mcp command. Use when the user asks to invoke, call, or execute an MCP tool, or when they need to send a JSON payload to an MCP server through the CLI. The JSON payload is forwarded 1:1 to the platform's MCP call-tool API.

Invoke connected Unique spaces/subagents through the unique-cli subagent command. Use when the workspace exposes connected-space tools and you need to delegate a question or task to one of those configured assistants.

Search the Unique AI Platform knowledge base using the unique-cli search command, with automatic per-turn citation tracking so cited facts render as clickable reference chips and `<sup>N</sup>` footnotes on the Unique platform. Use whenever the user asks to find, search, or query documents or content on Unique, including filtering by folder or metadata. Also covers `unique-cli read <cont_id>` for reading the full indexed text of a document when its content ID is already known. NOTE: This search uses combined vector + full-text indexing. Excel (.xlsx/.xls), CSV (.csv), and image files are NOT full-text indexed, so they will not appear in search results. To locate these file types, use the unique-cli-file-management skill instead (browse folders with `unique-cli ls` to find them by name).

Search the documents uploaded for the CURRENT task/row (e.g. an Agentic Table row's attached files) via the `unique-cli uploaded-search "<query>"` command, with the same per-turn citation tracking as `unique-cli search` so cited facts render as `<sup>N</sup>` footnotes and clickable reference chips on the Unique platform. ALWAYS use this skill when the user refers to documents they uploaded/attached to this task, or when you need facts from the task's own attached files. These uploaded files are scoped to the chat, NOT to a knowledge-base folder, so they will NEVER appear in `unique-cli search` results no matter the folder or metadata filters. Use `unique-cli search` for the knowledge base and this command for the task's uploaded documents; the two are complementary and citation numbering is shared across them within a turn.

Manage files and folders on the Unique AI Platform using the unique-cli command-line tool. Use when the user asks to upload, download, delete, rename, list, find, restore versions, list versions, look for, or organize files and folders on Unique, or to read / view / quote the text contents of a known file (optionally by page or page range, e.g. "what's on page 5?", "read pages 10-12"), or when working with scope IDs (scope_*) or content IDs (cont_*). IMPORTANT: When a user says they are "looking for a file" or wants to "find a file", they typically mean locating it within the Unique AI Platform knowledge base — not on the local filesystem. Use this skill to browse and list folders/files on Unique to help them find it. This is also the preferred approach for locating Excel (.xlsx/.xls), CSV (.csv), and image files, as these file types are not full-text indexed and cannot be found via vector/full-text search.

Reproduce ai-repo PR checks locally with Poe and CI scripts, including per-package typecheck and coverage behavior. Use when validating changes before push or when user asks which local commands match CI.

Create, update, list, and delete Unique Dynamic Frontend Spaces using the `unique-cli dynamic-frontend` command. Use when deploying a generated Dynamic Frontend ZIP, updating an existing Dynamic Frontend Space bundle, listing manageable Dynamic Frontend Spaces, deleting a deployed Dynamic Frontend Space, or when the user mentions Dynamic Frontend Space deployment through the CLI.

Backport fixes from main onto release/YYYY.WW in the ai repo with cherry-pick and rebase merge so release-please changelogs stay correct. Use when hotfixing a release branch, cherry-picking to release/*, or preparing a release backport PR.

Deploy a new LLM model via LiteLLM (GitOps) and/or Azure (Terraform + node backend), add it to the ai toolkit, and verify end-to-end. Use when rolling out a model to any environment, adding a model to the toolkit, troubleshooting "model not available" errors, or finding pricing/token limits/model ids for any provider.

AI repo versioning and changelogs via release-please — never edit CHANGELOG.md or pyproject version by hand. Use when the user asks about releases, version bumps, changelogs, or preparing a release.

Search the public web and crawl URLs via the Unique AI Platform's `unique-cli web-search` command, with automatic per-turn citation tracking so cited web facts render as clickable external reference chips and `<sup>N</sup>` footnotes on the Unique platform. ALWAYS use this skill when the user asks to search the public web, research a topic online, look up news/articles/regulations, fetch the contents of one or more URLs, or perform any "two-phase" web research (search → review snippets → crawl selected URLs for full content). Routes through the Unique AI Platform's `/web-search-api` endpoints, so the engine (Google, Brave, Tavily, Jina, Firecrawl, …) and crawler are resolved server-side from the current environment — you never need to manage engine API keys yourself. Use this instead of fabricating answers from training data whenever fresh / source-cited information is needed. Do NOT use this skill for the internal knowledge base (`unique-cli search`) or for arbitrary tool execution (`unique-cli mcp`).

Triage and fix Dependabot alerts and CodeQL findings in the ai repository. Use when the user asks to address security vulnerabilities, fix Dependabot alerts, resolve CodeQL issues, or do security maintenance.

Guide branch creation, commit messages, and titles so the repo’s semantic workflow is enforced

Tabular and numerical data analysis with descriptive statistics and insights. Use when the user provides data, tables, CSVs, or numbers and wants analysis.

Financial factsheet analysis with key metrics extraction and investment rationale

Professional email drafting with appropriate tone and structure. Use when the user wants to write, compose, or draft an email.

Contract risk analysis, obligation mapping, and clause review. Use when the user asks to review, analyze, or check a contract or legal document.

ALWAYS use this skill when the user asks to schedule, automate, set up recurring tasks, create a cron job, or anything involving timed / periodic execution. Do NOT use the local system crontab — scheduled tasks are stored on the Unique AI Platform and executed by a Kubernetes CronJob, not locally. Scheduled tasks always run on the current assistant (`$UNIQUE_ASSISTANT_ID`); never ask the user which assistant to use, neither in chat nor via `unique-cli elicit ask`. This skill covers creating, listing, updating, enabling, disabling, and deleting scheduled tasks via the unique-cli schedule command.

After pushing code or creating a PR, automatically wait for CI checks to complete, fix failures, address Bugbot findings, and re-push in a loop until the PR is green. Use when the user pushes to a PR, creates a PR, or asks to get a PR to green.

Implement comprehensive error handling for Python code paths to keep services resilient and user-friendly. Use when failures are currently silent or exceptions leak through.

Diagnose and fix CI failures without leaving your editor.

Ask clarifying questions before implementing to ensure Python requirements are understood. Use when a task lacks detail, dependencies are unclear, or multiple interpretations are possible.

Delete local git branches whose remotes are gone, after verifying via gh CLI that each was actually merged.

Add documentation pages to an existing MkDocs + Entangled site, explain how to serve locally, help adapt documentation

Handle pull requests and review comments using GitHub CLI (gh). Use when the user wants to check PR comments, address review feedback, reply with commit hashes, or work with pull requests.

Manage Python dependencies, lock files, and project environments using Poetry.

Proactively draft and create pull requests by inferring title, refs, and body from git context, then confirm and open via gh CLI.

Guide reviewers through systematic, constructive peer review of a pull request.

Guide authors through a structured self-review before requesting peer review.

Check if a pull request is too large and guide splitting it into focused PRs.

Apply Google-style docstrings and inline comments to Python .py files. Use when documenting new functions/classes/modules, incrementally documenting touched code, or flagging rename opportunities as TODOs.

Debug Python code using tracebacks, pdb, and structured logging.

Write focused, deterministic, well-documented pytest tests for AI-authored code.

Refactor Python code to improve quality while keeping functionality intact. Use when cleaning up complexity, removing duplication, or tightening structure.

Fix all Python lint violations and enforce consistent formatting in one command.

Navigate execution flow from an entry point, mapping call chains, external dependencies, and side effects.

Manage Python dependencies, lock files, and project environments using the uv package manager.