eshp--deep-dream
Maintenance tool to keep the ESHP memory graph updated and in sync with codebase changes.
Install
mkdir -p .claude/skills/eshp-deep-dream && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/15968" && unzip -o skill.zip -d .claude/skills/eshp-deep-dream && rm skill.zipInstalls to .claude/skills/eshp-deep-dream
Activation
This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.
Do deep maintenance on the eshp memory graph files, ensure the mem-graph is in sync with the codebaseKey capabilities
- →Perform deep maintenance on ESHP memory graph files
- →Ensure the memory graph is in sync with the codebase
- →List all relationships using `eshp rels`
- →List all tags using `eshp tags`
- →Summarise recently created and recalled memories using `eshp summarise`
- →Improve relationships by using specific terms
How it works
This skill performs maintenance on the ESHP memory graph by checking its accuracy against the codebase, consolidating tags and relationships, and adding missing entries.
Inputs & outputs
When to use eshp--deep-dream
- →Sync memory graph with codebase
- →Consolidate memory tags
- →Update project relationship graph
- →Refactor memory documentation
About this skill
ESHP Deep Dream
ESHP (electric sheep pen) is a codebase level memory graph for maintaining information about the current project.
This skill is for doing deep maintenance on the graph, ensuring it is up to date and that the relationships are mapped correctly.
Important: prefer eshp recall <slug> and eshp scan <topic> to reading
memory files directly.
Procedure
Ensure the git state is clean so we can commit the updated memory files.
Understand the current memory graph:
eshp rels- list all relationshipseshp tags- list all tagseshp summarise- list recently created and recalled memories
Understand the current codebase:
- High level goals
- Features and capabilities
- Developer guidance and lifecycle
- Any upcoming or planned features
- Issues, refactoring opportunities, etc
Ensure the memory graph is an accurate reflection of the project and covers planned and existing features, dev lifecycle, and crucial knowledge. Key question: if the code was removed could we recreate the project from the memory graph alone?
Maintain memory accuracy:
- use
eshp scan <query>andeshp recall <slug>to check for accuracy - use
eshp tagsand look for opportunities to consolidate similar tags - use
eshp relsand look for opportunities to consolidate similar relationships - add any missing entries
- add any missing tags or relationships
When writing or updating notes, follow the eshp--write-node skill.
Run eshp diagnose and treat the output as a work queue. Address issues in priority order:
- Dangling edges (broken links — fix immediately)
- Orphaned nodes (unconnected — link or delete)
- Notes with
##in body (format violation — split into separate notes) - Bloated notes (too long — split; the facts inside belong in separate nodes)
Once done commit the updated memory graph files as a new deep-dream commit.
Improve relationships
- Rels like 'related' and 'linked-to' don't provide much semantic meaning. Prefer specific relationships like 'caused', 'blocks', 'implements', 'waiting-for'.
- Do not use generic relationships if a more specific one exists. Replace vague names with precise ones as meaning becomes apparent.
When not to use it
- →When direct reading of memory files is preferred over `eshp recall` or `eshp scan`
- →When the git state is not clean
Limitations
- →Requires a clean git state to commit updated memory files
- →Prefers `eshp recall <slug>` and `eshp scan <topic>` over direct file reading
- →Focuses on improving semantic meaning of relationships
How it compares
This skill provides a structured procedure for maintaining a codebase-level memory graph, ensuring its accuracy and semantic richness, which differs from ad-hoc documentation updates.
Compared to similar skills
eshp--deep-dream side by side with the closest alternatives in the catalog.
| Skill | Installs | Updated | Safety | Difficulty |
|---|---|---|---|---|
| eshp--deep-dream (this skill) | 0 | 2mo | No flags | Intermediate |
| deepwiki-rs | 25 | 9mo | Review | Intermediate |
| repo-research-analyst | 1 | 6mo | Review | Intermediate |
| c4-component | 1 | 4mo | No flags | Intermediate |
Try saying
Example prompts that trigger this skill in your AI assistant.
You might also like
deepwiki-rs
sopaco
AI-powered Rust documentation generation engine for comprehensive codebase analysis, C4 architecture diagrams, and automated technical documentation. Use when Claude needs to analyze source code, understand software architecture, generate technical specs, or create professional documentation from any programming language.
repo-research-analyst
parcadei
Analyze repository structure, patterns, conventions, and documentation for understanding a new codebase
c4-component
sickn33
Expert C4 Component-level documentation specialist. Synthesizes C4 Code-level documentation into Component-level architecture, defining component boundaries, interfaces, and relationships. Creates component diagrams and documentation. Use when synthesizing code-level documentation into logical components.
agent-pseudocode
ruvnet
Agent skill for pseudocode - invoke with $agent-pseudocode
spec
matteocervelli
>
cs-arch
wildlily1021
维护 `codestable/architecture/` 这份只记现状的系统地图,三种模式 update / check / backfill。触发:用户说"刷新 architecture"、"做架构检查"、"补这个模块的架构文档"、"方案和代码对得上吗",或 feature 阶段需要先做架构动作。不写未来规划(走 cs-roadmap)。