Greywall's built-in pi profile leaves the agent's persistence vectors writable. Analysis of the gaps (network, inaccurate XDG paths, overly broad allowWrite) and a corrective denyWrite profile, with the limits the sandbox cannot cover on its own.
Ai
pi-secured-setup v1.1.0 maps against the OWASP AI Agent Security Cheat Sheet. HMAC-chained audit log, bash exfiltration detection, rate limiting, prompt injection scanner, and 14 Dependabot advisories closed.
The implementation of argentic-mw: the common Core in two hundred lines, three pipelines (Watch, Web Stats, Zendesk), the rule table, the auto-response guardrails and the scheduler via Postgres leader election.
Build or buy for an LLM agent? Three axes, six decision criteria, and a quadrant to compare full-LLM platforms, orchestration frameworks and a custom core. The framework I apply to argentic-mw.
Why I built a supervised agent rather than an autonomous one for a scheduled business pipeline. The LLM steps in at extension points where it is useful, the rest is deterministic, traceable, replayable code.
Example-based tests miss the edges. Property-based testing flips the approach. I show six useful patterns, stateful PBT for sequence bugs, and why this topic is well-suited for AI-assisted use.
On a 100% private local RAG, replacing nomic-embed-text with bge-m3 and dropping runtime FR→EN translation took recall from 0.45 to 0.95 and cut latency by twelve. Measured, not guessed.
A synthesis of Mario Zechner's ideas on coding agents. Why the minimalism of the Pi agent, coupled with strict context engineering and CLI tools, often beats heavy frameworks.
Two releases on the same day, twenty minutes apart: v1.0.1 then v1.0.2. Both do the same job — harden pi-secured-setup itself. This is a short post to explain what changed and why it matters, because a security tool that ships its own vulnerabilities loses all credibility.
If you’re new to the project, the introductory article covers the basics: Guards, Scanners, the audit trail.
v1.0.1 — the quality pass
Three threads:
Once the spec is in place, I split the work between humans, AI, and deterministic tools. The critical test does not always come from the same author as the code, and the critic-agent often matters more than the writer-agent.