Add resonance: Agent Skills — recovered and updated#77
Merged
Conversation
Originally drafted Mar 5-6 alongside 'When Skills Aren't Enough.' Lost to same handoff friction. Recovered Apr 9. Updated both sides: skills gained subagents, evals, resource folders, hot-reload. oddkit gained Vodka Architecture name, E0007 proactive posture. ODD Principle rewritten to align with Vodka Architecture. All seven divergence points refreshed against current state.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What
Canon resonance page comparing Agent Skills with oddkit. Originally drafted Mar 5-6, lost to handoff friction, recovered Apr 9.
Updates since original
Skills side: Subagent delegation, eval/benchmark pipeline, resource subfolders, hot-reloading, 2600+ skills ecosystem.
oddkit side: Vodka Architecture named, E0007 proactive posture realized, six design principles formalized.
ODD Principle: Renamed from "Knowledge Should Load Progressively" to "The Serving Layer Carries Everything and Tastes Like Nothing" — aligns with Vodka Architecture.
All seven divergences refreshed. Core thesis holds: oddkit is what skills become when the collection outgrows the recipe box and the content shifts from instructions to judgment.
Validation
Litmus test: mechanical alignment yes, explicit divergence yes, consequential divergence yes. Template structure followed. Challenge passed.
Note
Low Risk
Low risk documentation-only change that adds a new resonance article; no runtime behavior, APIs, or data handling is modified.
Overview
Adds a new resonance canon page,
canon/resonance/agent-skills.md, comparing Anthropic’s Agent Skills standard with oddkit.The document captures mechanical alignments (progressive disclosure, portability, discovery/content separation) and explicit divergences (search vs triggers, variable-depth retrieval, interconnected corpus, vodka architecture, epistemic tools, continuous posture, thin prompts), and links to related canon including
vodka-architectureand the companion essay.Reviewed by Cursor Bugbot for commit 389fd01. Bugbot is set up for automated code reviews on this repo. Configure here.