This video explains why large collections of copied agent skills consume context and trigger irrelevant workflows, then argues for a small set of understood, frequently used skills tailored to your own work. It also shows how project-specific skills and selective cross-model review can make agents more effective without loading every workflow globally.
David Ondrej42 minTranscript found
Quick learning frame
Read this before watching.
Agent ops treats agents like services: observable state, queues, permissions, logs, recovery, and post-run review.
New playlist item from David Ondrej; queued for transcript-backed review, topic mapping, and a practical learning artifact.
Skill you build: The ability to audit an agent-skill collection and keep only focused, useful workflows at the scope where they belong.
Watch for the shift from claim to mechanism. The learning value is the point where the transcript reveals a repeatable action, tool boundary, context move, review habit, or artifact.
Concept diagram
Where this video fits.
01Project state
02Session
03Queue/Kanban
04Tools
05Logs
06Recovery
07Post-run review
Deep lesson
Turn this video into working knowledge.
9,073 cleaned transcript words reviewed across 2,777 timed caption segments.
Thesis
More skills = worse agents (do this instead) teaches a practical hermes operations move: This video explains why large collections of copied agent skills consume context and trigger irrelevant workflows, then argues for a small set of understood, frequently used skills tailored to your own work. It also shows how project-specific skills and selective cross-model review can make agents more effective without loading every workflow globally.
The goal is not to remember the video. The goal is to extract the operating principle, tie it to timestamped evidence, test how far the claim transfers, and make something reusable.
1:05
Descriptions Still Cost
“some value, but the point is if you bloat your own directory with skills, you're going to be wasting a lot of tokens. You're going to be hitting your usage limits faster. You're going to be actually hurting...”
A skill can defer loading its full workflow until invocation, but every skill description still enters the agent's context; hundreds of broad descriptions create cumulative token waste and can cause entire skill bodies to load unnecessarily. List your installed skills, mark which descriptions are broad enough to trigger often, and identify the least-used 80 percent as deletion candidates.
16:41
Review Across Models
“You can use any harness, right? So whether that is Claude Code, Cursor, Codex, Py, Hermes, Open Code. We have loads of different ones. This is Kimi, this is Droid, both on local and on Cloud Threads. And...”
For a medium or large change, the speaker uses an OpenAI model to review work produced by an Anthropic model, reserving that second pass for meaningful or uncertain changes rather than trivial visual edits. Choose one substantial recent agent-generated change and have a frontier model from a different provider review it for concrete defects.
37:05
Scope Skills Locally
“And the reason for that is like when you work in a specific project, usually the workflows are specific to that project, right? It sounds obvious, but like even generic skills like code review, it's it can be...”
Global skills load in every session, while project-specific skills stay relevant to one repository and can encode that codebase's particular review or operating practices. In teams, an expert can commit those specialized skills so everyone working in the repository benefits from the expert's process. Move one workflow that applies to only a single repository out of your global skill set and rewrite it with that project's conventions.
01
Project state
Start with this video's job: This video explains why large collections of copied agent skills consume context and trigger irrelevant workflows, then argues for a small set of understood, frequently used skills tailored to your own work. It also shows how project-specific skills and selective cross-model review can make agents more effective without loading every workflow globally. Treat "Project state" as the outcome you are trying to make visible, not a topic label. Anchor it to 1:05, where the video says: “some value, but the point is if you bloat your own directory with skills, you're going to be wasting a lot of tokens. You're going to be hitting your usage limits faster. You're going to be actually hurting...”
02
Session
Use "Session" to locate the part of the hermes operations mechanism the video is demonstrating. Ask what changes in your real setup if this claim is true. Anchor it to 16:41, where the video says: “You can use any harness, right? So whether that is Claude Code, Cursor, Codex, Py, Hermes, Open Code. We have loads of different ones. This is Kimi, this is Droid, both on local and on Cloud Threads. And...”
03
Queue/Kanban
Turn "Queue/Kanban" into the reusable artifact for this lesson: A Hermes-style agent-ops runbook with health checks, state transitions, logs, recovery steps, and review criteria. This is where watching becomes something you can inspect and reuse.
04
Tools
Use "Tools" as the application surface. Decide whether the idea touches a browser flow, a local file, a model choice, a source document, a UI, or a review step.
05
Logs
Use "Logs" to prove the lesson. The evidence should connect back to the video title, transcript anchors, and a concrete output, not a generic best-practice claim.
06
Recovery
Use "Recovery" to carry the idea forward: save the prompt, checklist, diagram, or operating rule that would make the next agent run better.
07
Post-run review
Connect "Post-run review" to More skills = worse agents (do this instead) by naming the claim, the evidence, and the artifact it should produce.
Example
Source-backed artifact packet
Convert the video into a scoped artifact request that includes the transcript claim, mechanism, acceptance criteria, and proof. The output should be a hermes-style agent-ops runbook with health checks, state transitions, logs, recovery steps, and review criteria..
Example
Hermes operations proof brief
Separate what the speaker claims, what the demo actually proves, and what still needs outside verification before you adopt the hermes operations pattern.
Example
Teach-back module
Transform the lesson into a definition, a Project state -> Session -> Queue/Kanban -> Tools -> Logs -> Recovery -> Post-run review diagram, one misconception, one practice exercise, and a check-for-understanding question.
Do not learn it wrong
Treating the title as the lesson without checking what the transcript actually says.
treating UI features as reliability
missing logs
no stop/recover path
Letting the lesson drift into feature cheerleading.
Letting the lesson drift into ops advice without logs/state.
Letting the lesson drift into assuming reliability from a demo alone.
Do not count this as learned until these are true.
01
State the transcript-backed claim in your own words: This video explains why large collections of copied agent skills consume context and trigger irrelevant workflows, then argues for a small set of understood, frequently used skills tailored to your own work. It also shows how project-specific skills and selective cross-model review can make agents more effective without loading every workflow globally.
02
Explain the practical stakes without hype: New playlist item from David Ondrej; queued for transcript-backed review, topic mapping, and a practical learning artifact.
03
Map the idea onto the Brief -> Source material -> Generation -> Selection -> Edit -> Taste review -> Reusable recipe sequence and name the weakest link.
04
Produce the artifact and include the evidence that proves it: A creative production board with source inputs, prompt recipe, selection criteria, edit pass, and taste-review checkpoints.
Put it into practice
Give this grounded prompt to Codex or Claude after watching.
You are helping me turn one specific YouTube video into real, durable learning.
Source video:
- Title: More skills = worse agents (do this instead)
- URL: https://www.youtube.com/watch?v=45-qbcAeUbM
- Topic: Creative Automation
- My current learning frame: Audit your installed skills, remove or demote the ones you rarely use, and turn one recurring repository workflow into a narrowly described project-specific skill.
- Why this matters: New playlist item from David Ondrej; queued for transcript-backed review, topic mapping, and a practical learning artifact.
Transcript anchors from this exact video:
- 1:05 / Evidence 1: "some value, but the point is if you bloat your own directory with skills, you're going to be wasting a lot of tokens. You're going to be hitting your usage limits faster. You're going to be actually hurting..."
- 9:32 / Evidence 2: "once auditing my entire codebase. So, what you need to understand is your computer has dot agents slash skills. This is the standard that everybody follows. Even Cloud Code, you know, recently they changed it. For the longest..."
- 13:50 / Evidence 3: "even though I just use Opus 5.5 as my main model and I only use GPT for review. Yeah, OpenAI is not the best in terms of the usage limits right now, but that's a topic for a..."
- 16:41 / Evidence 4: "You can use any harness, right? So whether that is Claude Code, Cursor, Codex, Py, Hermes, Open Code. We have loads of different ones. This is Kimi, this is Droid, both on local and on Cloud Threads. And..."
- 33:45 / Evidence 5: "concept. Your skills should also work in Cloud Agents. A lot of people, the reason why they don't adopt Cloud is because their full setup is local, right? All their skills, MCPs, APIs, whatever, re- GitHub repos, projects,..."
- 37:05 / Evidence 6: "And the reason for that is like when you work in a specific project, usually the workflows are specific to that project, right? It sounds obvious, but like even generic skills like code review, it's it can be..."
- 41:56 / Evidence 7: "you don't have to switch between Codex, Grokbot, Claude desktop, right? That's a complete mess. You have all your work in a single place and uh as I mentioned, it's the only AI coding tool in the world..."
Video-aware target:
- Prompt lane: Hermes operations
- Mechanism to extract: Identify the operations control that makes long-running agent work visible, recoverable, or safer.
- Artifact to produce: A Hermes-style agent-ops runbook with health checks, state transitions, logs, recovery steps, and review criteria.
- Artifact must include: health check; state model; permission boundary; log source; recovery action
Your task:
1. Use the transcript anchors above as the primary source packet. If you add outside context, label it clearly as outside context and keep it secondary.
2. Create a source-check table with columns: timestamp, claim, transcript support, what the demo proves, confidence, and what still needs verification.
3. Extract the actual teachable mechanism from the video: Identify the operations control that makes long-running agent work visible, recoverable, or safer. Do not invent claims that are not supported by the title, lesson frame, or transcript anchors.
4. Build a reusable learning artifact: A Hermes-style agent-ops runbook with health checks, state transitions, logs, recovery steps, and review criteria.
5. Include:
- a plain-English definition of the core idea
- a diagram or structured model using this sequence: Project state -> Session -> Queue/Kanban -> Tools -> Logs -> Recovery -> Post-run review
- answers to these source questions: What operational failure is prevented? | What state is visible? | What can be recovered or redirected?
- 3 concrete examples that apply the video idea to real agentic work, such as Hermes Kanban triage; local model endpoint check; agent swarm recovery review
- 2 failure modes the video helps prevent, chosen from the transcript evidence and these likely risks: treating UI features as reliability; missing logs; no stop/recover path
- a checklist for the next real workflow, focused on: status, model/backend, tools, logs, recovery
- one practical exercise with a clear done signal: Write a runbook for restarting one stuck Hermes-style agent session.
6. Add a "learning transfer" section: what changes in my workflow tomorrow if I actually learned this?
7. Add a "source check" section that cites which transcript anchor supports each major takeaway.
Quality bar:
- Make this specific to "More skills = worse agents (do this instead)", not a generic Creative Automation essay.
- Ground each ops recommendation in transcript evidence about state, queues, models, tools, security, logs, or recovery.
- Prefer operational examples, failure modes, and reusable artifacts over broad definitions.
- Call out uncertainty instead of smoothing over weak evidence.
- Avoid these generic drifts: feature cheerleading; ops advice without logs/state; assuming reliability from a demo alone.
- If evidence is weak or missing, stop and say what transcript segment or timestamp needs review instead of guessing.
- Finish with a concise artifact I could paste into my learning app.
Misconceptions
What to stop believing.
Creative AI removes the need for taste.
It increases the need for taste because output volume explodes.
The best prompt is enough.
References, critique, iteration, and post-production matter just as much.
Practice studio
Learning only counts when you make something.
01
Transcript evidence map
Separate what the video actually says from what you already believe about the topic.
3 source-backed takeaways with timestamps, confidence, and a transfer note.02
One useful artifact
Apply the video to a real workflow and produce a hermes-style agent-ops runbook with health checks, state transitions, logs, recovery steps, and review criteria..
A reusable artifact with a done signal and one verification step.03
Hermes operations teach-back card
Explain the hermes operations mechanism to someone who has not watched the video yet.
A 90-second explanation, one diagram, one example, and one misconception to avoid.
Recall check
Answer first, then reveal — without rewatching.
Why can hundreds of skills waste context even when their full bodies load only on invocation?
When does the speaker recommend cross-model review rather than skipping it?
Why are project-specific skills often preferable to global skills?
Source shelf
Use the video as a doorway, then verify with primary sources.