Nice. Does this work across multiple machines / moved repos (same project, different local path), or is it strictly path-based?

it reads the session files that claude writes locally, so works on a single local machine only, and breaks if you move a repos around on the same machine.

The ASCII output is the missing piece for AI-assisted coding workflows. LLMs can spit out Mermaid, but you can't see the diagram inline in a terminal/code-review context. This fixes that.

When I needed to see mermaid diagrams rendered the other day I realized my existing Forgejo instance already supported it so that's what I've been doing. Push to repo and check it out there. Not inline but not a bad place for code reviews. (I am sure all the other forges also support it)

How did you catch it — scanner, review, or just noticed manually? I treat agent-generated diffs as untrusted by default now.

I was manually reviewing when I saw it. Was looking through the PR more out of interest than worried that there'd be a problem tbh.

How do you handle side-effectful tools and secrets? Least-privilege and auditing seem hard once any Python function can become a tool (scoped creds, revocation, logs).

Hi Thanks for the comment, The unified agent doesn’t execute untrusted code — it orchestrates. Side effects only happen at explicit execution boundaries (remote tools or the CodeAgent). The CodeAgent is sandboxed in Docker; the planner/router isn’t, because it never executes arbitrary code. Least-privilege and auditing are enforced at the execution boundary, not per function. Do you have any suggestions for making PolyMCP better?

Makes sense — execution boundary is the right place.

I'd prioritize audit logs + correlation IDs, and short-lived creds per tool call. Do you expose tool capabilities to the planner without exposing creds?

Orchestration buys parallelism, not coherence. More agents means more drift between assumptions. Past a point you're just generating merge conflicts with extra steps.

NVLink postRxDetLinkMask errors show up right before the hang. Has anyone captured a bug report or stack trace while nvidia-smi is stuck to see what it's blocking on?

Looks like an airplane, fails under real inspection. Same smell in diffs: bugfix turns into refactor. Review eats the "velocity." Revert rate keeps you honest.

Merge cost kills this. Does the harness enforce file/ownership boundaries per worker, and run tests before folding changes back into the lead context?

I don't know what you're referring to but I can say with confidence that I see more efficient token usage from a delegated approach, for the reasons I stated, provided that the tasks are correctly sized. ymmv of course :)

The “craftsman to Ikea factory manager” line from the interview is the real headline here. AI does the fun creative stuff, you get stuck reviewing 2000 lines you didn’t write. Revert rate tells you more than any “10x” claim.

Triage time is the real cost. Even obvious junk takes a couple minutes to read. Surprised bounty platforms dont have a "prove you ran it" step -- version, command, actual output.