Agents

Hermes Agent Challenge Submission: Write About Hermes Agent

Mukunda Rao Katta

May 25

My agent started lying. It took me two days to realize the system prompt had changed.

#devchallenge #hermesagentchallenge #agents #python

3 min read

Hermes Agent Challenge Submission: Build With Hermes Agent

Mukunda Rao Katta

May 25

Claude's extended thinking is useful. Sending raw <thinking> tags to users is not.

#devchallenge #hermesagentchallenge #agents #python

2 min read

Hermes Agent Challenge Submission: Build With Hermes Agent

Mukunda Rao Katta

May 25

I logged 300 Hermes runs to one file. trace-session-split cut it into 300.

#devchallenge #hermesagentchallenge #agents #observability

2 min read

Hermes Agent Challenge Submission: Build With Hermes Agent

Mukunda Rao Katta

May 25

One Hermes tool call took 4.8 seconds. The others averaged 900ms. trace-anomaly found it.

#devchallenge #hermesagentchallenge #agents #observability

3 min read

Hermes Agent Challenge Submission: Write About Hermes Agent

Mukunda Rao Katta

May 25

Your Hermes agent's audit log is leaking customer emails. Here's a 100-line lib that fixes that.

#devchallenge #hermesagentchallenge #agents #security

4 min read

Hermes Agent Challenge Submission: Build With Hermes Agent

Mukunda Rao Katta

May 25

My Hermes agent kept returning JSON in a markdown code block. I kept writing the same regex. Then I stopped.

#devchallenge #hermesagentchallenge #agents #python

2 min read

Hermes Agent Challenge Submission: Build With Hermes Agent

Mukunda Rao Katta

May 25

I replaced 200 lines of ad-hoc state management in my Hermes agent with one object.

#devchallenge #hermesagentchallenge #agents #python

2 min read

Hermes Agent Challenge Submission: Write About Hermes Agent

Mukunda Rao Katta

May 25

Your Hermes agent's prompt cache is 30% hit because of one timestamp. Here's how to find it.

#devchallenge #hermesagentchallenge #agents #performance

4 min read

Hermes Agent Challenge Submission: Build With Hermes Agent

Mukunda Rao Katta

May 25

My agent kept forgetting what it was doing. A scratchpad fixed it.

#devchallenge #hermesagentchallenge #agents #python

2 min read

Hermes Agent Challenge Submission: Build With Hermes Agent

Mukunda Rao Katta

May 25

I ran three Hermes agents in parallel. trace-merge stitched them into one timeline I could read.

#devchallenge #hermesagentchallenge #agents #observability

5 min read

Hermes Agent Challenge Submission: Write About Hermes Agent

Mukunda Rao Katta

May 25

The two-line Hermes agent logger I wish existed a month ago

#devchallenge #hermesagentchallenge #agents #observability

3 min read

Hermes Agent Challenge Submission: Build With Hermes Agent

Mukunda Rao Katta

May 25

My Hermes agent ran 500 turns and cost $40 before I noticed. Now it can't.

#devchallenge #hermesagentchallenge #agents #python

2 min read

Hermes Agent Challenge Submission: Write About Hermes Agent

Mukunda Rao Katta

May 25

My Hermes agent's stop condition was a 40-line if/elif chain. I replaced it with 3 lines.

#devchallenge #hermesagentchallenge #agents #python

2 min read

Hermes Agent Challenge Submission: Build With Hermes Agent

Mukunda Rao Katta

May 25

My Hermes agent called exec_shell. It shouldn't have been able to.

#devchallenge #hermesagentchallenge #agents #python

2 min read

Hermes Agent Challenge Submission: Write About Hermes Agent

Mukunda Rao Katta

May 25

My Hermes agent spent $3 before I noticed. Now it can't.

#devchallenge #hermesagentchallenge #agents #python

2 min read

DEV Community

# agents

My agent started lying. It took me two days to realize the system prompt had changed.

Claude's extended thinking is useful. Sending raw <thinking> tags to users is not.

I logged 300 Hermes runs to one file. trace-session-split cut it into 300.

One Hermes tool call took 4.8 seconds. The others averaged 900ms. trace-anomaly found it.

Your Hermes agent's audit log is leaking customer emails. Here's a 100-line lib that fixes that.

My Hermes agent kept returning JSON in a markdown code block. I kept writing the same regex. Then I stopped.

I replaced 200 lines of ad-hoc state management in my Hermes agent with one object.

Your Hermes agent's prompt cache is 30% hit because of one timestamp. Here's how to find it.

My agent kept forgetting what it was doing. A scratchpad fixed it.

I ran three Hermes agents in parallel. trace-merge stitched them into one timeline I could read.

The two-line Hermes agent logger I wish existed a month ago

My Hermes agent ran 500 turns and cost $40 before I noticed. Now it can't.

My Hermes agent's stop condition was a 40-line if/elif chain. I replaced it with 3 lines.

My Hermes agent called exec_shell. It shouldn't have been able to.

My Hermes agent spent $3 before I noticed. Now it can't.