Automation

TL;DR I have $X in monthly Claude tokens I don’t always use. Instead of letting the unused credit evaporate, I built a parallel agent sweep that fans out autonomous scouts to scan for dependency upgrades, CVEs, CI waste, and quick wins across my repos. Each discovery agent returns a scored candidate list. The orchestrator triages and ranks them, then spins up isolated worktree agents to implement the safe ones — all under a hard token cap and with human gates between phases. The output is a pile of merge requests, not silent commits. Noise is real and review burden is the limiting factor, but when it lands right, an hour of agent work + human review beats a weekend of manual maintenance. ...

TL;DR I’ve wired LLMs into my GitLab CI pipeline to auto-review merge requests, post findings as comments, and (on command) generate patches and commit fixes. The key insight: deterministic gates run first. Before the LLM ever sees a diff, regex-enforced checks block deleted tests, committed secrets, and destructive commands. Regex is certain; LLM judgment is probabilistic. Gate first, judge second. The bot reviews silently unless it finds something, posts to the MR with confidence levels, and can be leveled up from read-only observer to trusted committer as it proves itself — hence the “autonomy ladder” (Rungs 0–4) that gates who decides what. Infrastructure repos cap at Rung 2 (never auto-merge). ...

TL;DR I run unattended LLM agents on my homelab — they write code, open MRs, generate content, rotate secrets. The problem: they fail silently and bill silently. Langfuse (a tracing platform) logs every LLM call with input/output tokens, latency, and cost. On top of those traces, I built three background monitors that run weekly: a goal-drift detector that compares an agent’s stated objective to what its commits actually did (via embedding similarity), a cost-spike alert that fires at 80% and 100% of a daily budget cap, and an action audit that exports traces and flags sessions where the tool-call sequence diverged from the plan. Together, these let me sleep while autonomous agents handle repetitive work. ...

TL;DR Model Context Protocol (MCP) is a transport layer that lets Claude and other LLM agents call local tools with typed signatures and structured responses. Any HTTP API running on your homelab — ComfyUI, a wiki, a dashboard, a custom service — can become a set of agent-callable tools by wrapping it in a FastMCP server. A typical server takes 150–250 lines of Python, exposes 3–5 tools via @mcp.tool() decorators, and runs as a stdio process. The pattern scales from single-purpose (image generation) to multi-tool (queue status, model listing, system stats) without complexity explosion. This post shows the anatomy by dissecting the ComfyUI MCP server: how to build workflows, poll for completion, parse results, and return structured JSON that agents actually use. ...

TL;DR A new Opus model shipped, so I sat down to re-tune the agent harness I drive it with — the CLAUDE.md files, skills, hooks, and settings that shape every session. The surprising part: the most valuable changes weren’t trimming prompts for the smarter model. They were wiring the agent into infrastructure I already run — offloading bulk work to a local LLM (≈$0), a live homelab statusline, session tracing for an “action audit,” and a goal-drift monitor that uses the local model as judge. I also learned not to trust the new model’s own suggestions about what to cut. It wanted to delete load-bearing guardrails. ...

TL;DR I run a growing fleet of autonomous agents — homelab ops, trading research, content generation. Most blow up the first few times they try anything new. I needed a way to decide what an agent is allowed to do without asking me, and what still requires a human checkpoint. The answer is a four-rung trust ladder — supervised, monitored, trusted, full autonomy. Agents earn rungs through track record, not promises. Demotions are possible and routine. The framework took the question “should this agent be allowed to do X” out of my head every single time and turned it into a policy I can apply consistently. ...

TL;DR I built an “Ops Dream Worker” — a Kubernetes CronJob that runs at 3 AM, inspects the cluster, identifies improvements, and files GitHub issues with specific fixes. It runs entirely on local models (Mac Studio M3 Ultra), costs $0 per run, and went through 240 A/B test iterations to optimize the prompts. The anti-hallucination patterns were harder to get right than the analysis itself. The idea I have a k3s cluster with ~40 deployed services. I maintain it solo. There’s always something that could be better — a deployment missing resource limits, a CronJob that’s been failing silently, an ingress without SSO protection, a container image with known CVEs. These improvements pile up because I’m usually focused on building features, not auditing infrastructure. ...

TL;DR I built a complete swing trading research platform from an empty scaffold to 13,674 lines of Python in a single day. Five phases: data layer and backtesting, fundamentals and sentiment, portfolio construction, ML signals and Monte Carlo, then paper trading with a terminal dashboard. 199 tests across 48 test files. The platform fetches from Yahoo Finance, FRED, SEC EDGAR, and news APIs, runs technical and fundamental analysis, backtests strategies with walk-forward validation, and presents recommendations through a Rich terminal dashboard with human-in-the-loop approval. No cloud dependencies, no subscriptions, no vendor lock-in. ...

TL;DR A 109-day project plan. One day of actual work. Eight hours of active pipeline time. The key was treating planning and implementation as two separate AI-driven phases: spend an evening getting the plan right by routing it through multiple models, then let Claude Sonnet 4.6 implement it autonomously overnight via GitHub Copilot’s cloud agent while you sleep. This is the full playbook — planning phase included. The Project This came out of building dnd-multi, a full-stack AI Dungeon Master platform: FastAPI backend, Next.js 15 frontend, a Discord bot, LiveKit voice, and AWS Bedrock integration. Seven feature phases, a plan projected to take until June 19. ...

TL;DR A $20/month GitHub Copilot subscription gives you Claude Sonnet 4.6, GPT-4o, and Gemini inside VS Code. Out of the box it’s useful. With a proper instruction setup — a copilot-instructions.md file, path-scoped rules, and skill documents — it becomes something you actually rely on. Most of the posts on this blog were built with this toolchain, mostly in the context of my k3s cluster, but the patterns apply anywhere. This is how I have it set up. ...