ai-spector 0.3.4 → 0.3.6

package/scaffold/cursor/commands/generate-detail-design.md

@@ -1,19 +0,0 @@
-# /generate-detail-design
-
-Detail design via graph context. **User runs this command;** agent runs CLI. Graph-first: [_generate-graph.md](./_generate-graph.md). On CLI failure: [_cli-failures.md](./_cli-failures.md).
-
-## Prerequisites
-
-Graph validates; SRS minimum exists.
-
-## Required Behavior
-
-1. `ai-spector graph validate`
-2. Seed: `feature` id or detail `document` node.
-3. `ai-spector graph query <seed> --direction both --depth 3 --json`
-4. Load SRS + basic-design paths from `projectionPaths` only.
-5. Generate; update graph; `ai-spector graph validate`.
-
-## If blocked
-
-Failed validate/query → [_cli-failures.md](./_cli-failures.md). Do not read all of `docs/` manually when CLI failed.

package/scaffold/cursor/commands/generate-srs.md

@@ -1,183 +0,0 @@
-# /generate-srs
-
-Generate SRS markdown **from the traceability graph** — carefully, in DAG order, ingesting each file back into the graph.
-
-**User runs this command;** the agent runs CLI. Shared graph-first rules: [**_generate-graph.md**](./_generate-graph.md).
-
-## Philosophy
-
-- **Accuracy over speed** — full graph context before every write; ingest before the next wave.
-- **Graph in, graph out** — query neighbors + dependencies before; `rendersTo` + `dependsOn` after.
-- **Parallel when safe** — targets in the **same DAG wave** (no `dependsOn` between them) may be generated in one batch; never skip query/ingest per file.
-
-## Usage — three ways to choose targets
-
-| Case | User says | Agent behavior |
-|------|-----------|----------------|
-| **1 — All (default)** | `/generate-srs` | Plan **every** DAG node (respect `good` / skip unless regen requested). Full wave walk. |
-| **2 — Explicit files** | `/generate-srs docs/srs/3-use-cases.md` or `/generate-srs file1.md file2.md` | Map paths → DAG nodes → seeds. Add **DAG dependencies** not on disk yet. Batch only within same wave. |
-| **3 — Request in words** | `/generate-srs general use case chapter and features` | Resolve intent → proposed file list → **confirm with user** → then same as case 2. **Do not generate until user approves.** |
-
-Arguments after the command are either **file paths** (case 2) or a **free-text request** (case 3). If mixed, treat paths as explicit and text as request.
-
-## Case 3 — resolve request and confirm (mandatory)
-
-1. Parse the user’s words against `dag.srs.json`, `dag.srs.graph-seeds.json`, and graph domain nodes (`useCase`, `feature`).
-2. Build a **proposed scope** table:
-
-| # | DAG id | Output path | Seed | Reason matched | Also generate deps? |
-|---|--------|-------------|------|----------------|---------------------|
-| 1 | `srs.use-cases` | `docs/srs/3-use-cases.md` | `doc.srs.3-use-cases` | “use case chapter” | yes — `srs.introduction`, `srs.overall-description` if missing |
-
-3. Include **dependency closure**: any DAG `dependsOn` ancestor that is `missing_file` / `missing_content` must be listed (generated first in earlier waves) unless user says to skip deps.
-4. Ask explicitly:
-
-```text
-I plan to generate the following (N files, waves X–Y). Dependencies marked “prerequisite” run first.
-Reply **yes** to proceed, **no** to cancel, or edit the list (e.g. “skip introduction”, “add 5-data-requirements”).
-```
-
-5. **Stop** if the user does not confirm. Do not write SRS files on assumption.
-6. If the request is ambiguous (e.g. “features” = list only vs all F-xx detail files), ask one clarifying question **before** the confirmation table.
-
-### Intent → DAG hints (not exhaustive)
-
-| User phrase (examples) | Typical DAG / outputs |
-|------------------------|------------------------|
-| introduction, purpose, scope | `srs.introduction` |
-| overall, actors, users, §2 | `srs.overall-description` |
-| use case(s), UC, §3, chapter 3 | `srs.use-cases` (+ optional `UC-xx` detail if “all UC details”) |
-| feature list, §4.2, system features | `srs.features-list` |
-| feature detail, per feature, F-xx | `srs.feature-details` (`F-01`, … from graph) |
-| data, entities, §5 | `srs.data-requirements` |
-| interfaces, API, §6 | `srs.external-interfaces` |
-| NFR, quality, §7 | `srs.quality-attributes` |
-| i18n, localization, §8 | `srs.internationalization` |
-| legal, compliance, §9 | `srs.other-requirements` |
-| “everything” / no filter | Case 1 — full DAG |
-
-## Prerequisites
-
-`workflow.dependencies.json` → `generate-srs`. Requires merged graph (`/analyze`), passing **`ai-spector graph validate`**.
-
-Config:
-
-- `dag.srs.json` — generation order + `dependsOn`
-- `dag.srs.graph-seeds.json` — DAG id → `doc.srs.*` graph seed
-- `completeness-rules.srs.json` — quality checks
-
-## Required behavior
-
-### 1. Gate
-
-```bash
-ai-spector graph validate
-```
-
-Stop if errors — [_cli-failures.md](./_cli-failures.md).
-
-### 2. Plan (graph + DAG)
-
-1. **Select targets** using Usage case 1, 2, or 3 above. For case 3, run § “confirm” first.
-2. Topological sort **selected** nodes plus any required dependency ancestors.
-3. Map each node with `dag.srs.graph-seeds.json` → `targetSeedId` (`doc.srs.*` or `UC-xx` / `F-xx` for per-domain).
-4. Classify disk: `good` | `missing_content` | `missing_file` — do not overwrite `good` unless user asked to regenerate (explicit path or confirmed scope).
-5. Group into **waves** (see _generate-graph.md § Waves). Present final wave table (cases 1–2: brief; case 3: already confirmed).
-
-### 3. Per wave, then per target
-
-For **each wave** in order:
-
-1. All targets in this wave may run **in parallel** only if they do not depend on each other (true by wave definition).
-2. For **each** target in the wave (parallel or serial):
-
-#### 3a. Load context (CLI)
-
-Per [_generate-graph.md](./_generate-graph.md):
-
-- `graph query` each **DAG dependency** that should already exist (depth 2, include `rendersTo`).
-- `graph query` **target** seed (depth **4**, full generate edge set).
-- `graph impact <targetSeedId> --json` when replacing or ambiguous scope.
-
-Read **only** paths from `projectionPaths` plus targeted `docs/data-source/**` for gaps.
-
-#### 3b. Write one file
-
-- Read the DAG `template` file under **`.ai-spector/templates/`** (e.g. `.ai-spector/templates/srs/1-introduction.md`). Missing folder → user must run `npx ai-spector init`.
-- Content must match graph: UC list from `listedIn` on `useCase` nodes; features from `feature` + `satisfies`.
-- No fabricated requirements.
-
-#### 3c. Ingest into graph (mandatory before next wave)
-
-After each file **or** once per wave (combine all edges in one patch):
-
-```bash
-ai-spector graph merge .ai-spector/.docflow/extract/projection-patch.json
-ai-spector graph validate
-```
-
-Do not start the **next wave** until validate passes.
-
-Patch must include at minimum:
-
-- `rendersTo` — `from`: document id, `to`: actual file path
-- `dependsOn` — `from`: downstream doc id, `to`: each upstream doc id from DAG (same generation wave)
-
-Example for `srs.use-cases` → `doc.srs.3-use-cases`:
-
-```json
-{
-  "version": 1,
-  "nodes": [],
-  "edges": [
-    { "type": "rendersTo", "from": "doc.srs.3-use-cases", "to": "docs/srs/3-use-cases.md" },
-    { "type": "dependsOn", "from": "doc.srs.3-use-cases", "to": "doc.srs.1-introduction" },
-    { "type": "dependsOn", "from": "doc.srs.3-use-cases", "to": "doc.srs.2-overall-description" }
-  ]
-}
-```
-
-**Forbidden:** finishing without `rendersTo` on all generated documents; skipping merge between **waves**; batching targets from different waves (e.g. use-cases + features-list together).
-
-#### 3d. Log
-
-Append to `.ai-spector/.docflow/logs/generate-srs.log` (create if needed): timestamp, seed, path, validate OK.
-
-### 4. Finish
-
-- Final `ai-spector graph validate`.
-- **Refresh indexes and semantic graph** (parses UC/F/actor ids from generated markdown + updates Graphify on changed paths):
-
-```bash
-ai-spector index
-```
-
-`index` also adds **per-domain detail** `document` nodes (`doc.srs.uc-UC-01`, `doc.srs.f-F-01`, …) plus **`section` nodes** parsed from each detail file’s headings. **`definedIn`** links each UC/F to those sections (e.g. Overview, Main flow); **`rendersTo`** still points at the markdown path. Per-file `graph merge` patches are still required for list chapters (`3-use-cases.md`, `4-system-features.md`) and `dependsOn`; without them, validate may warn on missing `rendersTo` from template `doc.srs.*` nodes.
-
-Use `ai-spector index --force-graphify` if Graphify storage should rebuild even when file hashes are unchanged.
-- Optional: `/visualize-graph` for user review.
-- Suggest `/summary srs` only after graph ingest is complete.
-
-## Waves (reference)
-
-| Wave | DAG examples | Seed type |
-|------|----------------|-----------|
-| 0 | `srs.introduction`, `srs.overall-description` | `doc.srs.*` |
-| 1 | `srs.use-cases` | `doc.srs.3-use-cases` + domain queries for §3.2 |
-| 2 | `srs.features-list` | `doc.srs.4-system-features` |
-| 3 | `srs.feature-details` | `F-xx` per file |
-| 4 | data, interfaces, NFR chapters | `doc.srs.5-*` … with deps via `dependsOn` queries |
-
-## Guardrails
-
-- **Parallel only within a wave** — never across waves; never when A `dependsOn` B in the DAG.
-- **Every target** gets its own `graph query` + dependency queries before write (parallel OK).
-- **Every wave** ends with `graph merge` + `graph validate` before the next wave.
-- On `graph query` / `merge` / `validate` failure → stop per _cli-failures.md.
-- Prefer graph `nodes`/`edges` over `knowledge.json` for generation text.
-
-## If blocked
-
-[_cli-failures.md](./_cli-failures.md). User re-runs **`/generate-srs`** after fixes.
-
-Common fix: run `/analyze` if domain nodes missing; run ingest merge if downstream query returns empty `rendersTo` for dependencies.

package/scaffold/cursor/commands/graph-impact.md

@@ -1,5 +0,0 @@
-# /graph-impact (deprecated)
-
-Use **`/impact`** instead — see [impact.md](./impact.md).
-
-`/graph-impact <nodeId>` required graph ids. **`/impact`** resolves seeds from your selection, file paths, headings, and natural-language change descriptions.