Dynamic Skills

Why enigma's skills are dynamic - they adapt to the app that runs them, to your configuration, and to the task at hand, instead of being one fixed file.

A normal agent skill is static: one Markdown file, dropped into one tool, identical for everyone, changed only by editing the file. enigma’s policy skills are dynamic - the same source adapts to where it runs, to how you’ve configured enigma, and to what you’re doing.

Static vs dynamic

Static skill Dynamic skill (enigma)
Targets One tool, by hand Claude Code, Codex and opencode from one source
Per-app control None On/off per app
Reacts to your settings No Re-rendered from your enigma config
Loaded Always, in full By context (progressive disclosure)
Kept current Manual re-copy Auto-synced + updated from the repo

Adapts to the app

A skill is authored once and deployed into each agent in that agent’s own format and location - Claude Code, Codex and opencode each read skills and memory differently, and enigma handles the translation. You can keep a skill on for one app and off for another (enigma skills disable testing-policy opencode), and managed accounts inherit the right deployment automatically.

Adapts to your configuration

This is what makes them dynamic rather than just portable. The instructions enigma deploys are re-rendered from your settings, not copied verbatim. Turn on token-efficient output, change the minimal-code intensity, toggle commit-message emoji or parallel sub-agents - and the next sync rewrites the deployed memory and policy so the agent follows your chosen behavior. Change a setting, relaunch, and the guidance changes with it.

Loads by context

Only each skill’s name and one-line description sit in the agent’s context at all times. The full skill body loads when the task matches - via progressive disclosure - so a security task pulls in the security policy and a commit pulls in the git policy, without paying the token cost of every skill on every turn.

Stays current

Launching a tool through enigma (e.g. enigma claude) auto-syncs the deployed skills and memory to the latest version, and skills update straight from the repo between releases - every update verified by content hash. You install once; they keep themselves current.

See Configuration for the settings that drive the re-rendering, and Commands for enigma skills and enigma install.