pjt222/agent-almanac

Create, track, switch, sync, and clean up Git branches. Covers naming conventions, safe branch switching with stash, upstream synchronization, and pruning merged branches. Use when starting work on a new feature or bug fix, switching between tasks on different branches, keeping a feature branch up to date with main, or cleaning up branches after merging pull requests.

Complete procedure for submitting an R package to CRAN, including pre-submission checks (local, win-builder, R-hub), cran-comments.md preparation, URL and spell checking, and the submission itself. Covers first submissions and updates. Use when a package is ready for initial CRAN release, when submitting an updated version of an existing CRAN package, or when re-submitting after receiving CRAN reviewer feedback.

Consistent agent behavior after restart — progressive identity loading, working context reconstruction from persistent artifacts, fresh-vs-continuation detection, calibration through centering and attunement, and identity verification for coherence. Addresses the cold-start problem where an agent must reconstruct who it is and what it was doing from evidence rather than memory. Use at the start of every new session, after a session interruption or crash, when agent behavior feels inconsistent with prior sessions, or when persistent memory and current context appear contradictory.

Identify gemstones using optical properties, physical tests, and inclusion analysis. Covers refractive index, specific gravity, pleochroism, spectroscopy indicators, and common simulant detection. Use when identifying an unknown gemstone, verifying a seller's claim about species identity, distinguishing natural stones from simulants or synthetics, building gemological literacy through structured observation, or identifying rough material before cutting to ensure safe handling.

Create a new SKILL.md file following the Agent Skills open standard (agentskills.io). Covers frontmatter schema, section structure, writing effective procedures with Expected/On failure pairs, validation checklists, cross-referencing, and registry integration. Use when codifying a repeatable procedure for agents, adding a new capability to the skills library, converting a guide or runbook into agent-consumable format, or standardizing a workflow across projects or teams.