agent-environment-retr…
Use when a completed session needs an agent-environment retrospective. Not for an engineering retrospective from telemetry: use engineering-retrospective.
Use when a user resumes work and needs an orientation card without full human recall. Not for packaging context for another agent or session: use handoff. Never source or remote-system changes.
$ npx -y skills add OutlineDriven/odin-claude-plugin --skill gut-sync --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/gut-syncContext preview
The summary Claude sees to decide when to auto-load this skill.
Use when a user resumes work and needs an orientation card without full human recall. Not for packaging context for another agent or session: use handoff. Never source or remote-system changes.
name: gut-sync description: 'Use when a user resumes work and needs an orientation card without full human recall. Not for packaging context for another agent or session: use handoff. Never source or remote-system changes.' disable-model-invocation: true
| Field | Bound contract | |---|---| | Trigger | User resumes work and needs an orientation card without full human recall. | | Authority | Read-only. No file, VCS, credential, paid, published, deployed, or remote mutation. | | Side effect | Reads current workspace state, recent artifacts, and session context only. | | Done | A small orientation card is returned containing shape, invariants, and smell locations. |
1. Bound scope to read-only inspection of the current workspace directory and any session or handoff notes the user supplies. Do not open files outside that scope. Done when: scope is bounded to the supplied workspace directory and user-named notes, and no file outside that scope is opened. 2. List the workspace tree depth-limited and read the most recently modified files, any plan/todo/handoff notes, and reachable session context to recover what work was in flight. Done when: the workspace tree is listed depth-limited, the most recently modified files are read, and any plan/todo/handoff notes and session context are read or confirmed absent. 3. Derive the shape: name the active work unit, its current state, and the next intended action. Done when: the shape names the active work unit, its current state, and the next intended action, each sourced from workspace reads or marked unknown. 4. Derive the invariants: list constraints, contracts, or rules the work must not violate. Done when: the invariants list names every constraint, contract, or rule the work must not violate, each sourced from workspace reads or marked unknown. 5. Derive the smell locations: point at files or areas where decay, drift, risk, or unfinished work is concentrated. Done when: the smell locations point at specific files or areas where decay, drift, risk, or unfinished work is concentrated, each sourced from workspace reads or marked unknown. 6. Assemble the orientation card from steps 3-5 only. Mark any field that could not be recovered as unknown rather than guessing. Done when: the orientation card is assembled from the shape, invariants, and smell locations only, and every field that could not be recovered is marked unknown rather than guessed.
A small orientation card with three sections - Shape, Invariants, Smell locations - each populated from workspace reads or marked unknown.
Formerly the ODIN Claude Plugin. The repository URL is unchanged. Outline-Driven Development, nicknamed ODIN, is a highly opinionated code-agent skill library: principles-first engineering, surgical editing, and workflow automation, published as installable
Repo: OutlineDriven/odin-claude-plugin
Use when a completed session needs an agent-environment retrospective. Not for an engineering retrospective from telemetry: use engineering-retrospective.
Use when a repo needs agent setup, AGENTS.md added or made lean, CLAUDE.md audited, or agent instructions scored or pruned. Not for remote, credential,…
Use when a human explicitly asks for a full repository agent-compatibility pass returning a scored report with prioritized fixes. Not for tasks that require…
Use when setting up a project, auditing agent command permissions, or asking which read-only bash commands and domains to allow. Not for remote, credential,…
Use when asked to build or review a CLI intended for coding agents and return flag-driven, pipeline-safe, idempotent design advice. Not for running or…
Use when the user asks to make the skills framework work in a new harness, IDE, or CLI. Not for remote, credential, publish, deploy, or irreversible changes.