depslisted
Install: claude install-skill ivanviragine/wade
# Dependency Analysis
Analyze a set of GitHub issues and determine the dependency relationships
between them. Output a structured file that `wade` will use to generate
dependency graphs and update issue bodies.
## When to activate
- After `wade plan` creates multiple issues
- When `wade task deps` is run on existing issues
- When the user asks to analyze dependencies between issues
> **Note:** `wade task deps` first attempts headless analysis (AI tools that
> support `--print`/`--prompt`). If headless fails, it falls back to interactive
> mode: passes the analysis prompt directly to the AI tool as an initial message,
> then reads the output from a file after exit.
## Input
You will receive a context file containing one or more issues in this format:
```
=== Issue #41: Add user preferences schema ===
<issue body text>
=== Issue #42: Add preferences API endpoints ===
<issue body text>
=== Issue #43: Add preferences UI panel ===
<issue body text>
```
## Step 1: Analyze relationships
For each pair of issues, determine if there is a dependency. A dependency
exists when:
- One issue creates something (schema, module, API) that another issue uses
- One issue sets up infrastructure that another issue builds upon
- The implementation order matters for correctness (not just convenience)
**Do NOT create dependencies for:**
- Soft preferences ("it would be nice to do A first")
- Shared concerns without actual data/code flow
- Testing dependencies (tests can usually be writte