Read or write Airtable records — bases, tbl/rec/fld IDs, filterByFormula, pyairtable. Not for n8n workflows or Google Sheets/Excel.
Plan non-trivial features, fixes, refactors, UI/workflow changes before coding — clarify, research, get approval (Stop & Think protocol).
Find a past Claude Code session to resume — I closed my session, find the session where we…, which session discussed X, which session talked to the other agent.
Commit and optionally push. MANDATORY before any git write — commit, push, merge, rebase, reset, branch -d. Triggers: ship it, commit, push.
Review or improve an existing agent skill (SKILL.md structure, triggers, permissions, splitting). Not for new skills — use skills-build.
Build a NEW agent skill from a real repeated workflow. Use when the user asks to create a skill, scaffold `SKILL.md`, convert a recurring task into a skill, or decide whether a process deserves a skill. Do NOT use to review or audit an existing skill — use `skill-improver` for that.
Edit Tally forms via MCP tools — add/remove/reorder blocks and questions, change text, configure visibility/required state, and build conditional logic (apply_logic DSL). Use when the user mentions Tally, questionnaires, form building, form blocks, or a Tally form ID/URL, or asks to add a question, hide/show a field, sync Hebrew/English forms, or add logic to a Tally form. Does NOT trigger for writing the actual wording/persuasive copy of form questions or intro text (route that to `microcopy` or `copywriting`) — this skill is for structural/tool-mechanics edits via the Tally MCP, not for drafting prose.
Compare libraries/frameworks/tools and verify current APIs via live sources before choosing or implementing (X vs Y, latest, deprecated?). Also handles decision-grade questions — pivots, build-vs-buy, stack changes — via a weighted-criteria brief. Not for research already running inside a /design-log Phase B, and not for pure conceptual questions with no implementation impact.