Vincent Koc

Prepare and check trusted macOS hosts for physical Android and iPhone tests with agent-device, ADB, and Xcode, or set up iPhone Mirroring separately.

Resolve canonical Git checkouts, index and verify codebase-memory-mcp graphs through the guarded CLI, and safely audit duplicate worktree caches. Use when a user mentions codebase memory MCP, search_graph, trace_path, index_repository, worktree indexes, or oversized graph caches.

Mine structured Codex /goal history locally or across a configured machine fleet, measure active goal time and resumed thread spans, identify unfinished and recurring semantic runs, and turn them into stable copy-paste rerun suites. Use when the user asks to inspect goal history, summarize goal commands, find repeated goals, recover large beta campaigns, compare goal duration or token usage, or prepare reusable /goal prompts and privacy-scrubbed reports.

Recover Codex or Claude sessions from tmux panes, saved cockpit snapshots, session history, and recent agent logs without overwriting useful restore evidence. Use when a pane crashes, detaches, loses visible context, needs a resume command, or requires a safe targeted restore plan.

Interpret Crabpot and OpenClaw performance dashboard metrics, import-loop profiles, runtime profiles, and branch-to-branch perf deltas without over-reading noisy samples.

Maintain and release the crawlkit Go library, preserving downstream compatibility for gitcrawl, slacrawl, discrawl, and notcrawl.

Template skill for repository authors; excluded from public publishing.

Use when inspecting a ghcrawl SQLite store, pulling GitHub issue/PR data, refreshing summaries, embeddings, and clusters, or extracting one cluster and its evidence through the ghcrawl CLI.

Inspect or switch the local Git and GitHub download guard using an already-installed compatible low-data command. Use for airplane mode, flight or metered connections, returning online, or requests to enable, disable, or automatically select low-data protection on this machine.

Maintain, verify, and release graincrawl, the local-first Granola archive CLI, including SQLite archive behavior, read-only Granola source boundaries, Homebrew tap packaging, and crawlkit-powered TUI/snapshot surfaces.

Qualify and promote a pinned Bun fork build for a Linux OpenClaw service, including upstream sync, release proof, native workload admission, and rollback planning.

Investigate a cluster of GitHub issues and PRs, determine canonical candidates, post duplicate/related status, preserve contributor credit, and execute cleanup actions. Supports autonomous mode for provided-link-only closeout, merge/fix follow-through, changelog, and post-merge issue/PR cleanup.

Select, review, repair, validate, and land batches of up to 20 low-risk OpenClaw contributor pull requests using Vincent's maintainer preferences and bounded sub-agent lanes. Use for "next 20", broad contributor PR sweeps, merge-candidate mining, or continued PR-batch work where drafts, maintainer work, trivial one-line changes, UI, security, migrations, and high-risk changes must be excluded.

Create managed Git worktrees from a verified healthy owner, verify dependency reuse, and report explicit checkout and recovery outcomes.

Audit, compare, and document Opik integrations across Python SDK, TypeScript SDK, and OTEL/API backend. Use when adding a new integration, reviewing an existing one, or generating first-pass integration docs and gap analysis.

Optimize LLM prompts, tools, and agents in Opik using standardized optimizer workflows (prompt optimization, tool optimization, and parameter tuning), dataset/metric wiring, and result interpretation.

Audit and safely prune stale branches across a GitHub organization with immutable snapshots, conservative merged-PR classification, live SHA/protection/open-PR revalidation, resumable deletion ledgers, and post-delete verification. Use when a maintainer asks to clean up old, dead, merged, bot-created, or abandoned organization branches without risking default, protected, release, security-advisory, state, or active pull-request refs.

Build local semantic review slices by combining clawpatch feature maps, deepsec threat candidates, visual review maps, and optional gitcrawl/discrawl evidence for repos such as openclaw/openclaw.

Prepare a concise session handoff when the user asks to wrap up, capture continuation context, or use /done. Retain a Markdown file only when requested.

Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.

Remove AI-style noise from staged and branch diffs while preserving behavior and repository style.

Build, review, and audit technical docs and agent instruction files. Applies Simplified Technical English (ASD-STE100) to prose, scales to huge docs trees with sharded audits and ledgers, and runs as Claude Workflows or sub-agents when available.

Hunt existing integration patterns and design vendor/framework-agnostic API, RFC, SDK, and integration plans for new external vendor integrations.

Mine coding agent logs (Codex/Cursor/session histories and similar telemetry) to discover high-value candidate skills, then draft structured skill creation/reuse recommendations.

Monitor and coordinate one tmux agent lane, reconstruct worker state from panes and recent Codex logs, classify progress and blockers, and produce concise manager summaries. Use when multiple coding-agent workers run in tmux windows named L1, L2, and similar lane identifiers.

Manage one tmux agent lane from its matching ops pane, inspect live pane state and Codex logs on cold start, and produce concise manager summaries for OpenClaw and adjacent project work.