Deep Research Skill
A Claude skill that fans out parallel web research and synthesizes a cited report.
On the Air - LinkedIn Live, Every Friday 4:05 PM ET
Echo is the agent. Chris Shanku and Sash Mohapatra are the humans. Every Friday, Chris and Sash spend 15 minutes live-demoing real AI tools from their own systems. You see what works and what breaks. Echo brings it to life. That's Human in the Loop.
Real people. Real tools. Fifteen minutes.
Now Featured - Latest Drop
Also Worth Watching
Programming Schedule
Behind the Show
Every episode runs the same two-command pipeline. The infrastructure is the show.
One command creates the episode doc, generates show notes template, schedules the LinkedIn event, and stages the OBS overlay. Chris and Sash pick the tool - the pipeline handles everything else.
OBS to VDO.Ninja to LinkedIn RTMP. Real tool, real mistakes, real workarounds. The episode flips status automatically at 16:05 ET and marks aired at 16:20 ET - no manual intervention.
One command uploads to YouTube with chapters, publishes show notes, updates the episode archive, and posts the LinkedIn recap. Hosts talk - machines ship.
Echo knows the library. Ask her about any episode - tools covered, key moments, what was built. Every show is transcribed, summarized, and added to Echo's memory minutes after we wrap.
Ask Echo →Passing unit tests can still hide a broken system. 671 green tests checked the code, not whether the real automations fired - test the outcomes you actually depend on, not just the units.
EP010 - Watchdogs for Your Personal AITurn every incident into a permanent check. Log each break to a regressions.md and collect them into one smoke-test skill so the same failure can never silently return.
EP010 - Watchdogs for Your Personal AIKeep secrets out of chat logs - pipe credentials through a hook that injects from a vault into the environment, never into the prompt.
EP009 - Refactor Your AI, Don't Set ItTreat MCP as the API surface for agents - if a vendor or community MCP exists, try it before writing custom API plumbing.
EP008 - API vs MCPSplit your personal AI stack into a body (always-on, reachable) and a brain (heavy session-based work). Don't try to make one tool do both.
EP007 - My Agents Had a Convo at Midnight Without MeUse plan mode in Claude Code (Shift+Tab) before you let it write anything - it cannot touch files, so a misclick is harmless and the output is sharper.
EP006 - The 5-Step Method for Handing Real Work to AIA Claude skill that fans out parallel web research and synthesizes a cited report.
Drive tmux sessions from Claude Code - panes, windows, long-running jobs.
Interactive click-through of the regression + smoke-test watchdog setup.
How two agents coordinated overnight without a human in the loop.
The Builders
8 years at Microsoft FSI. Solution Engineer turned seller turned builder. Now Senior Director at iLink Digital, building AI systems that run in production every day. Created TheGrove - a personal AI operating system powered by Silas, with 23 agents and 41 skills.
20 years at Microsoft. Started as an Exchange server engineer in India, ended as Director of Data & AI Sales for Capital Markets. Now building what he wished existed during that transition - courses, newsletters, and AI-native products.
Every episode, Echo writes the copy, designs the branding, cuts the clips, and publishes the show notes before either host closes their laptop. The show's full transcript archive lives in a vector index Echo queries in real time. The only human approval gate is the LinkedIn post - everything else ships on its own.
Every Friday
EP.011 - Claude Code Minds OpenClaw airs June 12 at 4:05 PM ET. 15 minutes. One real tool. No safety net.