CorvinOS

Forge schema-bound, sandboxed tools at runtime. Paired with the cowork `forge` persona for chat-driven use in the voice-skill bridge. Use when you would otherwise write the same Bash snippet ≥3× with different parameters, when you need precise numerical results over a dataset, or when the user wants reproducible/deterministic output written to a file.

Voice mode for Claude Code — automatically reads assistant replies aloud via TTS, with German/English auto-detection and optional summarization for long responses. Trigger when the user mentions voice, vorlesen, sprachausgabe, TTS, "speak", "say it out loud", or asks about /voice-* commands.

ADR Gate — evaluates after non-trivial tasks whether an ADR is warranted; writes to Corvin-ADR repo when yes, names the skip reason when no, requires three-level analysis (conceptual/structural/implementation), and generates E2E tests for every new structural invariant so LDD has a loss signal.

Use when making any non-trivial recommendation, analysis, plan, design decision, code-review note, or architectural trade-off — before presenting to the user or acting. Forces a thesis / antithesis / synthesis pass so first-pass conclusions cannot ship without surviving their strongest counter-case.

Use when finishing any code change that modifies behavior, public API, CLI flags, config shape, defaults, error messages, or anything a reader-of-the-docs would model — before committing, pushing, or declaring "done." Forbids deferring doc updates to a follow-up commit, ticket, or "cleanup later."

Use periodically (weekly, at release candidates, before major version bumps) to scan a codebase for cumulative drift that no per-commit gate catches. Complements docs-as-definition-of-done, which prevents drift from being introduced per commit; this skill surfaces drift that already happened despite per-commit gates.

Use at the start of every inner-loop iteration where the goal is to fix a bug / close a failing test / reach a green E2E. Forbids editing code without first running the E2E to capture a fresh loss signal, and forbids declaring "done" without the E2E passing. Makes the cycle "E2E → loss → 5-why-by-layer → fix → E2E" the only admissible rhythm.

Use when a deliverable (document, diff, design, report, code module) is complete but "good enough, not great" and you want to improve it with a targeted re-iteration instead of starting over. Forbids re-running the task from scratch when refinement is more efficient. Operates on the y-axis (output) not the θ-axis (code or method).

Use at the start of any non-trivial engineering task (feature, bugfix touching more than one file, refactor with observable behavior change, incident response). Orchestrates the three LDD loops — inner (code), refinement (deliverable), outer (method) — with hard iteration budgets, dispatches the specialist skills at the right moments, and forbids declaring "done" without a synced doc-level mental model.

Use when deciding whether to edit code at all, how large the edit should be, or whether a "working" fix actually generalizes. Applies when iterating on a failing test/run, chasing a flake, committing a sequence of small patches, or tempted to ship on one data point. Frames code changes as gradient steps and forbids overfitting to a single sample or a single test.

Use when the same rubric violation, the same kind of symptom-patch, or the same escalation happens across 3+ distinct tasks — meaning the problem is in the method (skill, prompt, rubric), not in the tasks. Forces a disciplined outer-loop step: name the pattern, propose a skill/rubric change, measure Δloss before and after, roll back on regression.

Use when you have observed a failing test, a flaky run, a surprising log line, or any single-sample signal you are about to treat as a gradient. Forbids code edits based on one observation unless the log itself is unambiguous signal. Applies before invoking root-cause-by-layer or proposing any fix.

Use when encountering a bug, failing test, or unexpected behavior — especially under time pressure where try/except, retry loops, xfail, type-tolerance shims, "I'll clean up later," or compat-for-both-shapes feel like the fastest path. Forbids symptom patches until structural and conceptual origins are named.

Use at the start of any conversation, and whenever the user mentions LDD, loss-driven development, loss, gradient, SGD on code, drift, refinement loop, outer loop, inner loop, method evolution, or says "LDD:" / "apply LDD". Establishes how to dispatch the ten LDD skills and when each fires. Must be the first skill considered in any coding task where LDD is available.

Toggle quality disciplines (ADR Gate, docs-as-definition-of-done, etc.) on or off globally

Multi-persona layer for Claude Code. Per chat (or per CLI call) a different agent role: coder (default), browser, inbox, research, ... Each persona has its own tools, MCP servers, system prompt and workspace. Optional, on top of voice or standalone. Trigger when the user mentions cowork, persona, agent role, multi-agent, bind chat, /cowork-*.