Spec Driven Development, Workflows, and the Recent Coding Agent Conference
Get the full intelligence
Search transcripts, export clips, track mentions, and explore all topics from “Spec Driven Development, Workflows, and the Recent Coding Agent Conference” inside PodZeus.
This episode of MLOps.community dives deep into the insights and revelations from the recent Coding Agent Conference, exploring how AI agents are reshaping software development workflows. Hosts discuss the growing importance of agent architectures, sub-agents, adversarial agents, and the critical balance between specification-driven development and hands-on code review. A central theme is the shift from traditional coding to 'vibe coding'—AI-assisted development—where understanding the specification is paramount, but real-world debugging still demands direct code inspection. The hosts reflect on the conference's emphasis on orchestration, evaluation, and security, highlighting the lack of discussion around security risks like prompt injection and data exfiltration. They also explore practical tools and frameworks such as OpenClaw, Agent Zero, and PromptFoo, emphasizing the value of hooks for logging, monitoring, and evaluation. The conversation underscores a new paradigm in development: ownership, traceability, and the need for human oversight even in highly automated pipelines.
Specification-driven development is powerful when specs are precise, but a single typo can derail the entire project—always validate both spec and code.
Use 'learning tests' as early, codified spikes to validate ideas before full implementation, turning experimentation into reliable regression tests.
Adversarial agents and sub-agents are critical for robustness—use them to stress-test code, enforce guardrails, and improve quality.
Security is a major blind spot: prompt injection, unauthorized PRs, and API key leaks are real risks; use walled gardens, proxies, and sandboxed environments.
Hooks are a powerful, deterministic way to log agent behavior, track time, and evaluate performance—essential for observability in agentic workflows.
…and 2 more takeaways available in PodZeus
Conference Setup and First Impressions
The hosts set the scene for the episode by recalling their excitement about the Coding Agent Conference, including the surprise of guests traveling from afar like Germany and Seattle. They reflect on the team dinner the night before, filled with last-minute stress over slides and speaker readiness, setting the tone for a high-energy, chaotic yet inspiring event.
Post-Conference Data Processing and Skill Mining
“I took the transcript, fed it through Claude, and asked it to identify areas where we talked about certain skills. And if we weren't talking about existing skills, create that skill.”
The Rise of Agents and Sub-Agents
“The adversarial agent is just another agent that will be in that workflow checking the code as it's being created to try and poke holes in it.”
Evaluating Agents: The Black Box Problem
“There's no free lunch. Basically, you gain on one aspect, but you lose on another.”
Security and Trust in Agentic Workflows
“It was like every time, even in like the times where it like... gave him like three, you know, responses. You know, it's not blah, blah, blah. He got it to just, you know, do it.”
“It was like every time, even in like the times where it like... gave him like three, you know, responses. You know, it's not blah, blah, blah. He got it to just, you know, do it.”
“If your specification is so detailed that there's no possible way for the agent to derail from that plan, there's no reason to essentially read the code because your understanding of the code is less important than the specification that you put in place that it has now adhered to.”
“You own the code that your agent produces and you're the first line of defense. You have to be very confident that if you're going to submit that, that you've checked it over.”
Host
Guest
Claude
other
GitHub
other
OpenClaw
product
Dex
person
Matrix
other
Jess
person
Jesse
person
Agent Zero
product
PromptFoo
product
Warp
product
Fixing GPU Starvation in Large-Scale Distributed Training
MLOps.community • 52m • 4/3/2026
Getting Humans Out of the Way: How to Work with Teams of Agents
MLOps.community • 50m • 4/7/2026
We Cut LLM Latency by 70% in Production
MLOps.community • 1h 5m • 4/10/2026
The Modern Software Engineer
MLOps.community • 53m • 4/14/2026
Why Agents are Driving Software Development to the Cloud
MLOps.community • 51m • 4/17/2026
Get the full intelligence
Search transcripts, export clips, track mentions, and explore all topics from “Spec Driven Development, Workflows, and the Recent Coding Agent Conference” inside PodZeus.
Start discovering podcast insights today
Start with a 7-day trial and explore a growing catalog of popular podcasts. No credit card required.
No credit card required • 7-day trial • Cancel anytime