Skip to content

Commit 09101bb

Browse files
tirth8205claude
tirth8205
and
claude
committed
refactor(skill): split SKILL.md sections >500 lines into recipe files (#313)
Anthropic recommends SKILL.md <=500 lines (no guarantee that more is loaded into context). The /understand SKILL.md was 853 lines, well over the target. This change extracts large embedded bash/JS code blocks and JSON schema reference material into adjacent `recipes/` and `reference/` files referenced from SKILL.md. Extracted recipes (skills/understand/recipes/): - worktree-redirect.md (Phase 0 git worktree detection bash) - plugin-root-resolution.md (Phase 0 plugin discovery + build bash) - understandignore-generator.md (Phase 0.5 starter file generator) - incremental-update.md (Phase 2 incremental update branch) - inline-validator.md (Phase 6 default-path validator script) - llm-review-path.md (Phase 6 --review dispatch block) - fingerprints-baseline.md (Phase 7 fingerprint baseline) - intermediate-cleanup.md (Phase 7 cleanup bash) Extracted reference (skills/understand/reference/): - knowledge-graph-schema.md (node/edge type tables, weights) - output-shapes.md (layers[*], tour[*], assembled-graph shapes) SKILL.md line count: 853 -> 533. All workflow semantics preserved - the skill is still runnable end-to-end via the `see X` cross-references. Other SKILL.md files in the repo were already under 500 lines and were not touched. Updated comment in worktree-redirect.test.mjs to point at the new recipe location for the bash snippet (test content unchanged). Closes #313 Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
1 parent 7f5a717 commit 09101bb

11 files changed

Lines changed: 376 additions & 314 deletions

File tree

‎understand-anything-plugin/skills/understand/SKILL.md‎

Lines changed: 12 additions & 311 deletions
Large diffs are not rendered by default.

‎understand-anything-plugin/skills/understand/recipes/fingerprints-baseline.md‎

Lines changed: 26 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,26 @@
1+
# Recipe: Phase 7 Structural Fingerprints Baseline
2+
3+
> Builds the fingerprint baseline that future auto-update runs compare against. Referenced from `SKILL.md` Phase 7 step 2 (Generate structural fingerprints baseline).
4+
5+
Generate structural fingerprints baseline. This creates the basis for future automatic incremental updates and **must succeed before `meta.json` is written** — otherwise auto-update sees a fresh commit hash with no fingerprints to compare against, classifies every file as STRUCTURAL, and escalates to `FULL_UPDATE` on every subsequent commit (issue #152).
6+
7+
Write the input file:
8+
```bash
9+
cat > $PROJECT_ROOT/.understand-anything/intermediate/fingerprint-input.json <<EOF
10+
{
11+
"projectRoot": "$PROJECT_ROOT",
12+
"sourceFilePaths": [<all source file paths from Phase 1, as JSON array>],
13+
"gitCommitHash": "<current commit hash>"
14+
}
15+
EOF
16+
```
17+
18+
Then invoke the bundled script (located next to this SKILL.md):
19+
```bash
20+
node <SKILL_DIR>/build-fingerprints.mjs \
21+
$PROJECT_ROOT/.understand-anything/intermediate/fingerprint-input.json
22+
```
23+
24+
The script uses `TreeSitterPlugin + PluginRegistry` exactly like `extract-structure.mjs`, so the baseline matches the comparison logic used during auto-updates.
25+
26+
**If the script exits non-zero or stdout does not include `Fingerprints baseline:`, abort Phase 7 and report the error. Do NOT proceed to write `meta.json`.**

‎understand-anything-plugin/skills/understand/recipes/incremental-update.md‎

Lines changed: 27 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,27 @@
1+
# Recipe: Phase 2 Incremental Update Path
2+
3+
> The "Incremental update path" branch of Phase 2 — how to re-batch only changed files, prune the prior graph, and re-merge. Referenced from `SKILL.md` Phase 2 → "Incremental update path".
4+
5+
Write the changed-files list (one path per line) to a temp file:
6+
```bash
7+
git diff <lastCommitHash>..HEAD --name-only > $PROJECT_ROOT/.understand-anything/tmp/changed-files.txt
8+
```
9+
10+
Run compute-batches with `--changed-files`:
11+
```bash
12+
node <SKILL_DIR>/compute-batches.mjs $PROJECT_ROOT \
13+
--changed-files=$PROJECT_ROOT/.understand-anything/tmp/changed-files.txt
14+
```
15+
16+
This produces a `batches.json` that contains only batches with changed files, but neighborMap entries still reference unchanged files (with their full-graph batchIndex) so cross-batch edges remain emittable.
17+
18+
Then dispatch file-analyzer subagents per the same template as the full path.
19+
20+
After batches complete:
21+
1. Remove old nodes whose `filePath` matches any changed file from the existing graph
22+
2. Remove old edges whose `source` or `target` references a removed node
23+
3. Write the pruned existing nodes/edges as `batch-existing.json` in the intermediate directory
24+
4. Run the same merge script — it will combine `batch-existing.json` with the fresh `batch-*.json` files:
25+
```bash
26+
python <SKILL_DIR>/merge-batch-graphs.py $PROJECT_ROOT
27+
```

‎understand-anything-plugin/skills/understand/recipes/inline-validator.md‎

Lines changed: 80 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,80 @@
1+
# Recipe: Inline Deterministic Graph Validator
2+
3+
> The Node.js script written to `$PROJECT_ROOT/.understand-anything/tmp/ua-inline-validate.cjs` for the default (non-`--review`) Phase 6 path. Referenced from `SKILL.md` Phase 6 → "Default path (no `--review`): inline deterministic validation".
4+
5+
Write the following Node.js script to `$PROJECT_ROOT/.understand-anything/tmp/ua-inline-validate.cjs`:
6+
7+
```javascript
8+
#!/usr/bin/env node
9+
const fs = require('fs');
10+
const graphPath = process.argv[2];
11+
const outputPath = process.argv[3];
12+
try {
13+
const graph = JSON.parse(fs.readFileSync(graphPath, 'utf8'));
14+
const issues = [], warnings = [];
15+
if (!Array.isArray(graph.nodes)) { issues.push('graph.nodes is missing or not an array'); graph.nodes = []; }
16+
if (!Array.isArray(graph.edges)) { issues.push('graph.edges is missing or not an array'); graph.edges = []; }
17+
const nodeIds = new Set();
18+
const seen = new Map();
19+
graph.nodes.forEach((n, i) => {
20+
if (!n.id) { issues.push(`Node[${i}] missing id`); return; }
21+
if (!n.type) issues.push(`Node[${i}] '${n.id}' missing type`);
22+
if (!n.name) issues.push(`Node[${i}] '${n.id}' missing name`);
23+
if (!n.summary) issues.push(`Node[${i}] '${n.id}' missing summary`);
24+
if (!n.tags || !n.tags.length) issues.push(`Node[${i}] '${n.id}' missing tags`);
25+
if (seen.has(n.id)) issues.push(`Duplicate node ID '${n.id}' at indices ${seen.get(n.id)} and ${i}`);
26+
else seen.set(n.id, i);
27+
nodeIds.add(n.id);
28+
});
29+
graph.edges.forEach((e, i) => {
30+
if (!nodeIds.has(e.source)) issues.push(`Edge[${i}] source '${e.source}' not found`);
31+
if (!nodeIds.has(e.target)) issues.push(`Edge[${i}] target '${e.target}' not found`);
32+
});
33+
const fileLevelTypes = new Set(['file', 'config', 'document', 'service', 'pipeline', 'table', 'schema', 'resource', 'endpoint']);
34+
const fileNodes = graph.nodes.filter(n => fileLevelTypes.has(n.type)).map(n => n.id);
35+
const assigned = new Map();
36+
if (!Array.isArray(graph.layers)) { if (graph.layers) warnings.push('graph.layers is not an array'); graph.layers = []; }
37+
if (!Array.isArray(graph.tour)) { if (graph.tour) warnings.push('graph.tour is not an array'); graph.tour = []; }
38+
graph.layers.forEach(layer => {
39+
(layer.nodeIds || []).forEach(id => {
40+
if (!nodeIds.has(id)) issues.push(`Layer '${layer.id}' refs missing node '${id}'`);
41+
if (assigned.has(id)) issues.push(`Node '${id}' appears in multiple layers`);
42+
assigned.set(id, layer.id);
43+
});
44+
});
45+
fileNodes.forEach(id => {
46+
if (!assigned.has(id)) issues.push(`File node '${id}' not in any layer`);
47+
});
48+
graph.tour.forEach((step, i) => {
49+
(step.nodeIds || []).forEach(id => {
50+
if (!nodeIds.has(id)) issues.push(`Tour step[${i}] refs missing node '${id}'`);
51+
});
52+
});
53+
const withEdges = new Set([
54+
...graph.edges.map(e => e.source),
55+
...graph.edges.map(e => e.target)
56+
]);
57+
graph.nodes.forEach(n => {
58+
if (!withEdges.has(n.id)) warnings.push(`Node '${n.id}' has no edges (orphan)`);
59+
});
60+
const stats = {
61+
totalNodes: graph.nodes.length,
62+
totalEdges: graph.edges.length,
63+
totalLayers: graph.layers.length,
64+
tourSteps: graph.tour.length,
65+
nodeTypes: graph.nodes.reduce((a, n) => { a[n.type] = (a[n.type]||0)+1; return a; }, {}),
66+
edgeTypes: graph.edges.reduce((a, e) => { a[e.type] = (a[e.type]||0)+1; return a; }, {})
67+
};
68+
fs.writeFileSync(outputPath, JSON.stringify({ issues, warnings, stats }, null, 2));
69+
process.exit(0);
70+
} catch (err) { process.stderr.write(err.message + '\n'); process.exit(1); }
71+
```
72+
73+
Execute it:
74+
```bash
75+
node $PROJECT_ROOT/.understand-anything/tmp/ua-inline-validate.cjs \
76+
"$PROJECT_ROOT/.understand-anything/intermediate/assembled-graph.json" \
77+
"$PROJECT_ROOT/.understand-anything/intermediate/review.json"
78+
```
79+
80+
If the script exits non-zero, read stderr, fix the script, and retry once.

‎understand-anything-plugin/skills/understand/recipes/intermediate-cleanup.md‎

Lines changed: 20 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,20 @@
1+
# Recipe: Phase 7 Intermediate Cleanup
2+
3+
> Cleans up `.understand-anything/intermediate/` while preserving `scan-result.json`. Referenced from `SKILL.md` Phase 7 step 4.
4+
5+
Clean up intermediate files, **preserving `scan-result.json`** so future incremental runs can skip Phase 1 SCAN (see issue #293). We `mv` scratch dirs into a timestamped `.trash-*` instead of `rm -rf`ing them directly — this avoids tripping destructive-action gates on hardened hosts (e.g. freshness-window checks) that flag deleting directories created moments earlier (see issue #301). The delayed-purge step in Phase 0 reclaims the space once the trash is older than 7 days.
6+
7+
```bash
8+
# Preserve scan-result.json — Phase 1's deterministic file inventory.
9+
# Future incremental runs (Phase 2 compute-batches.mjs --changed-files=…)
10+
# need this inventory; without it, Phase 1 must re-dispatch and pay ~157k
11+
# tokens / ~158s per incremental run.
12+
TRASH="$PROJECT_ROOT/.understand-anything/.trash-$(date +%s)"
13+
mkdir -p "$TRASH"
14+
INTER="$PROJECT_ROOT/.understand-anything/intermediate"
15+
if [ -d "$INTER" ]; then
16+
# Move every entry except scan-result.json into the trash dir.
17+
find "$INTER" -mindepth 1 -maxdepth 1 -not -name 'scan-result.json' -exec mv {} "$TRASH/" \; 2>/dev/null || true
18+
fi
19+
mv "$PROJECT_ROOT/.understand-anything/tmp" "$TRASH/" 2>/dev/null || true
20+
```

‎understand-anything-plugin/skills/understand/recipes/llm-review-path.md‎

Lines changed: 26 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,26 @@
1+
# Recipe: Phase 6 `--review` Path (LLM Graph Reviewer)
2+
3+
> The alternative Phase 6 validation branch when `--review` is in `$ARGUMENTS`. Dispatches the LLM `graph-reviewer` subagent instead of the inline deterministic validator. Referenced from `SKILL.md` Phase 6 → "`--review` path: full LLM reviewer".
4+
5+
If `--review` IS in `$ARGUMENTS`, dispatch the LLM graph-reviewer subagent as follows:
6+
7+
Dispatch a subagent using the `graph-reviewer` agent definition (at `agents/graph-reviewer.md`). Append the following additional context:
8+
9+
> **Additional context from main session:**
10+
>
11+
> Phase 1 scan results (file inventory):
12+
> ```json
13+
> [list of {path, sizeLines} from scan-result.json]
14+
> ```
15+
>
16+
> Phase warnings/errors accumulated during analysis:
17+
> - [list any batch failures, skipped files, or warnings from Phases 2-5]
18+
>
19+
> Cross-validate: every file in the scan inventory should have a corresponding node in the graph (node types may vary: `file:`, `config:`, `document:`, `service:`, `pipeline:`, `table:`, `schema:`, `resource:`, `endpoint:`). Flag any missing files. Also flag any graph nodes whose `filePath` doesn't appear in the scan inventory.
20+
21+
Pass these parameters in the dispatch prompt:
22+
23+
> Validate the knowledge graph at `$PROJECT_ROOT/.understand-anything/intermediate/assembled-graph.json`.
24+
> Project root: `$PROJECT_ROOT`
25+
> Read the file and validate it for completeness and correctness.
26+
> Write output to: `$PROJECT_ROOT/.understand-anything/intermediate/review.json`

‎understand-anything-plugin/skills/understand/recipes/plugin-root-resolution.md‎

Lines changed: 51 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,51 @@
1+
# Recipe: Plugin Root Resolution and Build Check
2+
3+
> Resolves the `understand-anything` plugin root across install locations, then ensures `@understand-anything/core` is built. Referenced from `SKILL.md` Phase 0 step 1.5 (Ensure the plugin is built).
4+
5+
**Important:** do **not** assume the plugin root is simply two directories above the skill path string. In many installations `~/.agents/skills/understand` is a symlink into the real plugin checkout. Prefer runtime-provided plugin roots first (for Claude), then fall back to universal symlinks, skill symlink resolution, and common clone-based install paths.
6+
7+
Resolve the plugin root like this:
8+
9+
```bash
10+
SKILL_REAL=$(realpath ~/.agents/skills/understand 2>/dev/null || readlink -f ~/.agents/skills/understand 2>/dev/null || echo "")
11+
SELF_RELATIVE=$([ -n "$SKILL_REAL" ] && cd "$SKILL_REAL/../.." 2>/dev/null && pwd || echo "")
12+
COPILOT_SKILL_REAL=$(realpath ~/.copilot/skills/understand 2>/dev/null || readlink -f ~/.copilot/skills/understand 2>/dev/null || echo "")
13+
COPILOT_SELF_RELATIVE=$([ -n "$COPILOT_SKILL_REAL" ] && cd "$COPILOT_SKILL_REAL/../.." 2>/dev/null && pwd || echo "")
14+
15+
PLUGIN_ROOT=""
16+
for candidate in \
17+
"${CLAUDE_PLUGIN_ROOT}" \
18+
"$HOME/.understand-anything-plugin" \
19+
"$SELF_RELATIVE" \
20+
"$COPILOT_SELF_RELATIVE" \
21+
"$HOME/.codex/understand-anything/understand-anything-plugin" \
22+
"$HOME/.opencode/understand-anything/understand-anything-plugin" \
23+
"$HOME/.pi/understand-anything/understand-anything-plugin" \
24+
"$HOME/understand-anything/understand-anything-plugin"; do
25+
if [ -n "$candidate" ] && [ -f "$candidate/package.json" ] && [ -f "$candidate/pnpm-workspace.yaml" ]; then
26+
PLUGIN_ROOT="$candidate"
27+
break
28+
fi
29+
done
30+
31+
if [ -z "$PLUGIN_ROOT" ]; then
32+
echo "Error: Cannot find the understand-anything plugin root."
33+
echo "Checked:"
34+
echo " - ${CLAUDE_PLUGIN_ROOT:-<unset CLAUDE_PLUGIN_ROOT>}"
35+
echo " - $HOME/.understand-anything-plugin"
36+
echo " - ${SELF_RELATIVE:-<unresolved path derived from ~/.agents/skills/understand>}"
37+
echo " - ${COPILOT_SELF_RELATIVE:-<unresolved path derived from ~/.copilot/skills/understand>}"
38+
echo " - $HOME/.codex/understand-anything/understand-anything-plugin"
39+
echo " - $HOME/.opencode/understand-anything/understand-anything-plugin"
40+
echo " - $HOME/.pi/understand-anything/understand-anything-plugin"
41+
echo " - $HOME/understand-anything/understand-anything-plugin"
42+
echo "Make sure the plugin is installed correctly."
43+
exit 1
44+
fi
45+
46+
if [ ! -f "$PLUGIN_ROOT/packages/core/dist/index.js" ]; then
47+
cd "$PLUGIN_ROOT" && (pnpm install --frozen-lockfile 2>/dev/null || pnpm install) && pnpm --filter @understand-anything/core build
48+
fi
49+
```
50+
51+
If `pnpm` is missing, report to the user: "Install Node.js ≥ 22 and pnpm ≥ 10, then re-run `/understand`."

‎understand-anything-plugin/skills/understand/recipes/worktree-redirect.md‎

Lines changed: 25 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,25 @@
1+
# Recipe: Worktree Redirect
2+
3+
> Detects whether `PROJECT_ROOT` is a git worktree and redirects output to the main repo root. Referenced from `SKILL.md` Phase 0 step 1 (Resolve `PROJECT_ROOT`).
4+
5+
If `PROJECT_ROOT` is inside a git worktree (not the main checkout), redirect output to the main repository root. Worktrees managed by Claude Code are ephemeral — `.understand-anything/` written there is destroyed when the session ends, taking the knowledge graph with it (issue #133). Detect a worktree by comparing `git rev-parse --git-dir` against `git rev-parse --git-common-dir`; in a normal checkout or submodule they resolve to the same path, in a worktree they differ and the parent of `--git-common-dir` is the main repo root.
6+
7+
```bash
8+
COMMON_DIR=$(git -C "$PROJECT_ROOT" rev-parse --git-common-dir 2>/dev/null)
9+
GIT_DIR=$(git -C "$PROJECT_ROOT" rev-parse --git-dir 2>/dev/null)
10+
if [ -n "$COMMON_DIR" ] && [ -n "$GIT_DIR" ]; then
11+
COMMON_ABS=$(cd "$PROJECT_ROOT" && cd "$COMMON_DIR" 2>/dev/null && pwd -P)
12+
GIT_ABS=$(cd "$PROJECT_ROOT" && cd "$GIT_DIR" 2>/dev/null && pwd -P)
13+
if [ -n "$COMMON_ABS" ] && [ "$COMMON_ABS" != "$GIT_ABS" ]; then
14+
MAIN_ROOT=$(dirname "$COMMON_ABS")
15+
if [ -d "$MAIN_ROOT" ] && [ "${UNDERSTAND_NO_WORKTREE_REDIRECT:-0}" != "1" ]; then
16+
echo "[understand] Detected git worktree at $PROJECT_ROOT"
17+
echo "[understand] Redirecting output to main repo root: $MAIN_ROOT"
18+
echo "[understand] (Set UNDERSTAND_NO_WORKTREE_REDIRECT=1 to keep PROJECT_ROOT as the worktree.)"
19+
PROJECT_ROOT="$MAIN_ROOT"
20+
fi
21+
fi
22+
fi
23+
```
24+
25+
Set `UNDERSTAND_NO_WORKTREE_REDIRECT=1` if you intentionally want a per-worktree graph (rare — most users want the redirect).

‎understand-anything-plugin/skills/understand/reference/knowledge-graph-schema.md‎

Lines changed: 45 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,45 @@
1+
# Reference: KnowledgeGraph Schema
2+
3+
> Authoritative node-type, edge-type, and edge-weight tables for the assembled `knowledge-graph.json`. Referenced from `SKILL.md` "Reference: KnowledgeGraph Schema".
4+
5+
## Node Types (13 total)
6+
7+
| Type | Description | ID Convention |
8+
|---|---|---|
9+
| `file` | Source code file | `file:<relative-path>` |
10+
| `function` | Function or method | `function:<relative-path>:<name>` |
11+
| `class` | Class, interface, or type | `class:<relative-path>:<name>` |
12+
| `module` | Logical module or package | `module:<name>` |
13+
| `concept` | Abstract concept or pattern | `concept:<name>` |
14+
| `config` | Configuration file (YAML, JSON, TOML, env) | `config:<relative-path>` |
15+
| `document` | Documentation file (Markdown, RST, TXT) | `document:<relative-path>` |
16+
| `service` | Deployable service definition (Dockerfile, K8s) | `service:<relative-path>` |
17+
| `table` | Database table or migration | `table:<relative-path>:<table-name>` |
18+
| `endpoint` | API endpoint or route definition | `endpoint:<relative-path>:<endpoint-name>` |
19+
| `pipeline` | CI/CD pipeline configuration | `pipeline:<relative-path>` |
20+
| `schema` | Schema definition (GraphQL, Protobuf, Prisma) | `schema:<relative-path>` |
21+
| `resource` | Infrastructure resource (Terraform, CloudFormation) | `resource:<relative-path>` |
22+
23+
## Edge Types (26 total)
24+
25+
| Category | Types |
26+
|---|---|
27+
| Structural | `imports`, `exports`, `contains`, `inherits`, `implements` |
28+
| Behavioral | `calls`, `subscribes`, `publishes`, `middleware` |
29+
| Data flow | `reads_from`, `writes_to`, `transforms`, `validates` |
30+
| Dependencies | `depends_on`, `tested_by`, `configures` |
31+
| Semantic | `related`, `similar_to` |
32+
| Infrastructure | `deploys`, `serves`, `provisions`, `triggers` |
33+
| Schema/Data | `migrates`, `documents`, `routes`, `defines_schema` |
34+
35+
## Edge Weight Conventions
36+
37+
| Edge Type | Weight |
38+
|---|---|
39+
| `contains` | 1.0 |
40+
| `inherits`, `implements` | 0.9 |
41+
| `calls`, `exports`, `defines_schema` | 0.8 |
42+
| `imports`, `deploys`, `migrates` | 0.7 |
43+
| `depends_on`, `configures`, `triggers` | 0.6 |
44+
| `tested_by`, `documents`, `provisions`, `serves`, `routes` | 0.5 |
45+
| All others | 0.5 (default) |

‎understand-anything-plugin/skills/understand/reference/output-shapes.md‎

Lines changed: 61 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,61 @@
1+
# Reference: Phase Output Shapes
2+
3+
> Canonical JSON shapes for `layers[*]`, `tour[*]`, and the assembled `KnowledgeGraph` object that the workflow normalizes to. Referenced from `SKILL.md` Phase 4, Phase 5, and Phase 6.
4+
5+
## `layers[*]` shape (Phase 4)
6+
7+
Each element of the final `layers` array MUST have this shape. All four fields (`id`, `name`, `description`, `nodeIds`) are required.
8+
9+
```json
10+
[
11+
{
12+
"id": "layer:<kebab-case-name>",
13+
"name": "<layer name>",
14+
"description": "<what belongs in this layer>",
15+
"nodeIds": ["file:src/App.tsx", "config:tsconfig.json", "document:README.md"]
16+
}
17+
]
18+
```
19+
20+
## `tour[*]` shape (Phase 5)
21+
22+
Each element of the final `tour` array MUST have this shape. Required fields: `order`, `title`, `description`, `nodeIds`. Preserve optional `languageLesson` when present.
23+
24+
```json
25+
[
26+
{
27+
"order": 1,
28+
"title": "Project Overview",
29+
"description": "Start with the README to understand the project's purpose and architecture.",
30+
"nodeIds": ["document:README.md"]
31+
},
32+
{
33+
"order": 2,
34+
"title": "Application Entry Point",
35+
"description": "This step explains how the frontend boots and mounts.",
36+
"nodeIds": ["file:src/main.tsx", "file:src/App.tsx"]
37+
}
38+
]
39+
```
40+
41+
## Assembled `KnowledgeGraph` shape (Phase 6)
42+
43+
The full top-level shape written to `knowledge-graph.json`:
44+
45+
```json
46+
{
47+
"version": "1.0.0",
48+
"project": {
49+
"name": "<projectName>",
50+
"languages": ["<languages>"],
51+
"frameworks": ["<frameworks>"],
52+
"description": "<projectDescription>",
53+
"analyzedAt": "<ISO 8601 timestamp>",
54+
"gitCommitHash": "<commit hash from Phase 0>"
55+
},
56+
"nodes": [<all nodes from assembled-graph.json after Phase 3 review>],
57+
"edges": [<all edges from assembled-graph.json after Phase 3 review>],
58+
"layers": [<layers from Phase 4>],
59+
"tour": [<steps from Phase 5>]
60+
}
61+
```

0 commit comments

Comments
 (0)