I build automation CLIs and AI-agent tooling — sharp single-purpose command-line tools, plus the production patterns (approval gates, schema validation, audit logs) that let agents touch real systems safely.
If you script things, run agents, or want CLIs that do one job well with JSON output you can pipe — this profile is for you. Follow for new CLIs and production-AI patterns as they ship.
New: merged into tetherto/qvac — the official repo of Tether's QVAC AI platform: #3729 · validate RAG queries before logging
(Aug 2026) · all contributions ↓
- capcut-cli — CLI to edit CapCut / JianYing drafts (subtitles, timing, speed, templates, cut long-form → shorts). No API; reads
draft_content.jsondirectly - draftcat — governed AI pipelines for service businesses: deterministic-first, operator-approved, single Go binary (MIT)
- agentic-task-system — your task manager is the best agent memory you're not using: hybrid retrieval (RRF) over TickTick / Obsidian
- agent-approval-gate —
draft → validate → approve → dispatch → auditpattern with JSON schemas, n8n workflow, and email-approval example - skillgate — deterministic finish-line gates for AI coding agents: a model-independent evaluator that blocks commit/publish until your definition-of-done passes. opencode, Claude Code, pre-commit, CI
- browserground — local UI-grounding specialist for hybrid AI agents: Qwen3-VL-2B LoRA, screenshot + target → strict JSON bbox. Drop-in for Claude Code, Codex, browser-use
Production AI Automation Notes — a running set of repos and gists on:
- approval gates for AI agents
- MCP server security (permissions, tenants, audit logs)
- Claude Code policies for company repos
- n8n workflows with human approval
- audit-log schemas
Follow if you're building agents that need to work outside demos.
| Gist | Topic |
|---|---|
| Validate before you log: the RAG query guard that belongs above your first log line (PAAN #17) | RAG query validation / TypeError Cannot read properties of null reading 'substring' / validate before logging / log-ordering test / whitespace-only query / log injection / zod guard / capped log preview (August 2026) |
| GPT-5 / o-series vision calls failing in Go: MaxCompletionTokens and the empty finish_reason=length response | gpt-5 vision empty response / this model is not supported MaxTokens please use MaxCompletionTokens / finish_reason length / go-openai ReasoningValidator / o-series temperature beta-limitations / reasoning tokens / max_completion_tokens (August 2026) |
| Retrieval that degrades instead of failing: keyword fallback when the embedder dies (PAAN #16) | agent memory degraded retrieval / Premature close / APIConnectionError / embedding provider down / keyword fallback / reciprocal rank fusion / RRF / per-source status / no vector database (August 2026) |
| Cross-source agent memory for Claude Code: GitHub + Notion + your task app, one RRF query | cross-source retrieval / fuse GitHub Notion TickTick / one ranked list / composite adapter / RRF / no vector database / MCP / Claude Code (June 2026) |
| GitHub issues as agent memory for Claude Code — hybrid retrieval, no vector DB | github mcp / github issues as agent memory / issues + discussions / fine-grained PAT / hybrid RRF retrieval / no vector database / Claude Code (June 2026) |
| Project | PR | What |
|---|---|---|
| tetherto/qvac | #3729  |
Validate RAG queries before logging them |
| Tencent/WeKnora | #2614 |
GPT-5 / o-series vision model support (max_completion_tokens) |
| Tencent/WeKnora | #835 |
Parallel tool calling support |
| steveyegge/beads | #2884 |
Multi-project support, Notion sync, backup/restore |
| e2b-dev/infra | #2273 |
Local dev docs: prerequisites, verification steps, troubleshooting |
| pacifio/cersei | #10 |
Native Google Gemini provider + Cohere & SambaNova support |
Stack: Python, Go, TypeScript, Node.js, Kubernetes, Linux, systemd, vector databases, LLM APIs.
Website: renezander.com · YouTube: @determa
A gist series on production-grade patterns for AI pipelines. Each entry covers one pattern with copy-pasteable code.
| # | Topic | Pattern |
|---|---|---|
| #1 | Agent Approval Gates | Schema-validated proposals, human review, deterministic dispatch |
| #2 | Token Budgets | Per-step, per-pipeline, per-day enforcement |
| #3 | Agentic Knowledge Base | Karpathy-style LLM wiki with pluggable adapters |
| #4 | CapCut / JianYing from an LLM agent | Deterministic JSON command boundary |
| #5 | SQLite Dedup + Crash Safety | WAL mode, seen_items, audit log |
| #6 | Prompt-Injection Defense | Input sanitization, schema validation |
| #7 | PDF Cite Verification | Auditable LLM extraction with per-fragment bounding boxes |
| #10 | Deterministic Step Pipelines | Fixed typed steps; the LLM never picks the next action |
| #11 | Pipeline Fixture Testing | Dry-run pipelines from JSON fixtures; zero API calls, deterministic CI |
| #12 | LLM Skills as YAML | Prompt + output_schema + role in versioned YAML; validated by a linter |
| #13 | Inbound Agent Webhook Auth | Constant-time bearer token, fail-closed on empty secret, async 202 dispatch |
| #14 | Self-Improving Voice Agent | Human-approved prompt diffs; harvest > group > propose > 2 approval gates > git commit > staging smoke > auto-versioned publish |
| #15 | AI Action Audit Trail | Append-only action_approvals table: who approved which payload, when; gate-violation query; GDPR Art. 22 provenance |
| #16 | Degraded Retrieval | Keyword fallback + RRF when the embedding provider dies; per-source status, degraded: true |
| #17 | Validate Before You Log | Guard above the first log line; capped log preview; an ordering test that isn't vacuous |
Reference implementation for entries #1, #2, #5, #6, #7, #9, #10, #11, #12, #13: draftcat (Go, MIT).