@codedrifters/configulator 0.0.202 → 0.0.203

package/docs/label-taxonomy.md

@@ -39,7 +39,7 @@ Owned by the individual agent bundle that depends on them. A bundle
 should contribute only the labels its rules reference — typically a
 small family of phase labels used by a multi-phase workflow.
 
-Examples seen in the vortex reference taxonomy:
+Examples seen in a reference taxonomy from a downstream project:
 
 - `meeting:extract`, `meeting:notes`, `meeting:draft`, `meeting:link`
   — owned by the meeting-analysis bundle
@@ -78,8 +78,8 @@ description by supplying the same label name via
 ### Tier 3 — Consumer-project labels
 
 Owned by the end project. These are domain-specific categories
-configulator cannot reasonably predict. Examples from the vortex
-reference project:
+configulator cannot reasonably predict. Examples from a reference
+downstream project:
 
 - `area:requirements`, `area:business-strategy`, `area:references`,
   `area:product`, `area:architecture` — area labels tied to a
@@ -133,6 +133,6 @@ belongs to.
 ## See also
 
 - `docs/audits/label-audit-308.md` — initial audit of configulator
-  labels against the vortex reference taxonomy
+  labels against a reference taxonomy from a downstream project
 - `src/workflows/sync-labels.ts` — source of truth for the Tier 1
   default labels

package/lib/index.js

@@ -1093,6 +1093,89 @@ var jestBundle = {
   }
 };
 
+// src/agent/bundles/project-context.ts
+var PROJECT_CONTEXT_PATH = "docs/project-context.md";
+var PROJECT_CONTEXT_READER_SECTION = [
+  "## Project Context",
+  "",
+  `Before doing any work, read \`${PROJECT_CONTEXT_PATH}\` at the`,
+  'repository root. It is the canonical answer to "what is this project',
+  'about?" \u2014 mission, domain vocabulary, in/out-of-scope capabilities, and',
+  "key stakeholders. Use it to frame judgment calls in this session.",
+  "",
+  `If \`${PROJECT_CONTEXT_PATH}\` does not exist, proceed with the current`,
+  "task and note the absence in your session log \u2014 the meeting-analyst and",
+  "requirements-analyst agents seed the file on their next run.",
+  "",
+  "You are a **read-only consumer** of this file. Do not edit it.",
+  "",
+  "---",
+  ""
+];
+var PROJECT_CONTEXT_MAINTAINER_SECTION = [
+  "## Project Context",
+  "",
+  `Before starting any phase, read \`${PROJECT_CONTEXT_PATH}\` at the`,
+  'repository root. It is the canonical answer to "what is this project',
+  'about?" \u2014 mission, domain vocabulary, in/out-of-scope capabilities, and',
+  "key stakeholders. Use it to judge relevance when scanning source",
+  "material in this session.",
+  "",
+  "### Seed on first use",
+  "",
+  `If \`${PROJECT_CONTEXT_PATH}\` does not exist, create it from this`,
+  "template and commit it alongside this phase's other outputs:",
+  "",
+  "```markdown",
+  "# Project Context",
+  "",
+  "> Canonical description of this project. Read by all agents before",
+  "> work. Updated by meeting-analyst and requirements-analyst when new",
+  "> scope, vocabulary, or stakeholders emerge.",
+  "",
+  "## Mission",
+  "TODO: one or two sentences on what this project exists to do.",
+  "",
+  "## Target Users",
+  "TODO: who uses this, in what role.",
+  "",
+  "## In-Scope Capabilities",
+  "TODO: bullet list of the capabilities this project owns.",
+  "",
+  "## Out-of-Scope",
+  "TODO: capabilities or concerns that are explicitly not this project's.",
+  "",
+  "## Domain Vocabulary",
+  "TODO: short glossary of domain terms, acronyms, and their definitions.",
+  "",
+  "## Key Stakeholders",
+  "TODO: named people or teams and what they care about.",
+  "",
+  "## References",
+  "TODO: links to BCM docs, competitive analysis, product roadmap, and",
+  "other authoritative sources.",
+  "```",
+  "",
+  "Fill whatever you can infer from the source material you are already",
+  "reading in this phase; leave `TODO:` placeholders for the rest.",
+  "",
+  "### Update on new facts",
+  "",
+  "When the source material you process reveals new project-scope",
+  "information \u2014 a new capability, a vocabulary term, an entity, a",
+  "stakeholder, an in/out-of-scope decision, or a shift in mission \u2014",
+  `append or revise the relevant section in \`${PROJECT_CONTEXT_PATH}\``,
+  "before closing the phase. Commit those edits with the phase's other",
+  "outputs.",
+  "",
+  "Keep edits surgical: short bullet additions, brief clarifications, or",
+  "single-line vocabulary entries. Treat the file as an accreting",
+  "reference, not a document you reshape every session.",
+  "",
+  "---",
+  ""
+];
+
 // src/agent/bundles/meeting-analysis.ts
 var meetingAnalystSubAgent = {
   name: "meeting-analyst",
@@ -1134,6 +1217,7 @@ var meetingAnalystSubAgent = {
     "",
     "---",
     "",
+    ...PROJECT_CONTEXT_MAINTAINER_SECTION,
     "## Traceability",
     "",
     "All outputs must be bi-directionally linked so any artifact can be traced",
@@ -1783,6 +1867,7 @@ var orchestratorSubAgent = {
     "",
     "---",
     "",
+    ...PROJECT_CONTEXT_READER_SECTION,
     "## Phase A: Startup",
     "",
     "```bash",
@@ -1949,6 +2034,9 @@ var issueWorkerSubAgent = {
     "doubt, treat the session as interactive \u2014 pausing for approval is always",
     "safe; committing without approval is not.",
     "",
+    "---",
+    "",
+    ...PROJECT_CONTEXT_READER_SECTION,
     "## Phase 1: Select an Issue",
     "",
     "If an issue number was provided in your instructions, use that issue.",
@@ -2298,6 +2386,7 @@ var prReviewerSubAgent = {
     "",
     "---",
     "",
+    ...PROJECT_CONTEXT_READER_SECTION,
     "## Phase 1: Identify the PR",
     "",
     "If a PR number was provided in your instructions, use that. Otherwise stop",
@@ -2953,6 +3042,9 @@ var requirementsAnalystSubAgent = {
     "Follow your project's shared agent conventions (`AGENTS.md`,",
     "`CLAUDE.md`, or equivalent) for all commit, branch, and PR rules.",
     "",
+    "---",
+    "",
+    ...PROJECT_CONTEXT_MAINTAINER_SECTION,
     "## Design Principles",
     "",
     "1. **Discover, don't write.** This agent identifies *what requirements are",
@@ -3053,7 +3145,7 @@ var requirementsAnalystSubAgent = {
     "",
     "| Scope | What to read | What to look for |",
     "|-------|-------------|-----------------|",
-    "| **BCM model doc** | One `{PREFIX}-NNN` doc under `<BCM_DOCS_ROOT>` | `## Vortex Relevance` (or equivalent project-relevance) section \u2014 gaps where capabilities exist but no FR/BR/INT addresses them |",
+    "| **BCM model doc** | One `{PREFIX}-NNN` doc under `<BCM_DOCS_ROOT>` | The doc's project-relevance section (commonly `## <Project> Relevance` or `## Strategic Implications`) \u2014 gaps where capabilities exist but no FR/BR/INT addresses them. Use `docs/project-context.md` to judge what is relevant. |",
     "| **Competitive analysis** | One `comp-*.md` doc under `<COMPETITIVE_ROOT>` | Feature comparison gaps \u2014 competitor features the product lacks requirements for |",
     "| **Product roadmap** | `<PRODUCT_ROOT>/prioritized-feature-roadmap.md` | Roadmap items without corresponding FRs |",
     "| **Entity taxonomy** | `<PRODUCT_ROOT>/entity-taxonomy.md` | Entities without CRUD requirements (FR), data requirements (DR), or security requirements (SEC) |",
@@ -3246,9 +3338,9 @@ var requirementsAnalystSubAgent = {
     "   ```",
     "",
     "3. **Update source documents.** In each BCM model doc or competitive",
-    "   analysis that was scanned, add a note in the `## Vortex Relevance` or",
-    "   `## Strategic Implications` section indicating that a requirement issue",
-    "   was created:",
+    "   analysis that was scanned, add a note in the project-relevance /",
+    "   strategic-implications section (whichever heading the source doc uses)",
+    "   indicating that a requirement issue was created:",
     "",
     "   ```markdown",
     "   - Gap addressed: see [<PREFIX>-<NNN>](<relative path to requirement doc>)",
@@ -3266,7 +3358,7 @@ var requirementsAnalystSubAgent = {
     "",
     "| Direction | Agent | What |",
     "|-----------|-------|------|",
-    "| Upstream | BCM Writer | Scans `## Vortex Relevance` sections for capability gaps |",
+    "| Upstream | BCM Writer | Scans capability-model docs for project-relevance gaps (judged against `docs/project-context.md`) |",
     "| Upstream | Company Research | Scans competitive analysis for feature comparison gaps |",
     "| Upstream | Meeting Analyst | Scans meeting extracts for requirement proposals |",
     "| Downstream | Requirements Writer (BCM Writer) | Creates `type:requirement` issues for the skill to draft |",