@gluecharm-lab/easyspecs-cli 0.0.3

package/resources/opencode-agents/agent-md-feature-detail.md

@@ -0,0 +1,95 @@
+# Agent: Feature detail (markdown)
+
+| Field | Value |
+| ----- | ----- |
+| **AGENT_ID** | `ctx-md-feature-detail` |
+| **Display name** | Feature detail |
+| **SRS-8** | §4.3 `FE-<nn>-<slug>.md`; §6.2 (parallel after `features-list.json`); **R7**, **AC5**, **R21** |
+| **Output** | `<worktree>/.gluecharm/context/FE-<nn>-<slug>.md` |
+| **Input (coordination)** | `<worktree>/.gluecharm/context/features-list.json` (repo-relative: `.gluecharm/context/features-list.json`) |
+| **Pattern** | §3.5.1 — Markdown detail |
+
+## Responsibility
+
+**`.opencode/` exclusion:** The worktree may contain **`.opencode/`** (materialized OpenCode agents, schemas, and tooling). It is **not** part of the analyzed product codebase. **Do not** use it to infer application behavior. **Never** cite paths under **`.opencode/`** in **`sourceReferences`**, UI/backend evidence arrays, Evidence index bullets, **`evidenceRefs`**, or any other code-grounding output.
+
+**Functional and technical** documentation for **one feature** whose **`code`** and **`slug`** come from **`.gluecharm/context/features-list.json`**. Feeds human-readable depth; structure for the index comes from JSON lists. That row **must** include **`sourceReferences`** (`minItems: 1`); seed the **Evidence index** with those path+line ranges (then add more from deeper reading). In list JSON, **`sourceReferences[].path`** is always a **file**, never a folder.
+
+**Spine context:** Non-trivial features **typically** map to **multiple** use cases in **`FE-<nn>-use-cases-list.json`**. In **Summary** or **Scope**, avoid implying this feature is a single monolithic journey when the codebase supports several; you may briefly point readers at sibling UCs by code/name when helpful.
+
+**Coordination JSON** lives **only** under **`.gluecharm/context/`**, not the worktree root. Read that path first; if missing, **glob** `.gluecharm/context/**/features-list.json` before failing.
+
+## Task (for `{{TASK_DESCRIPTION}}`)
+
+Document feature **{{FE_CODE}}** (name/slug from list): scope, behaviour, main dependencies, entry points, and links to code. Ground all claims in the worktree.
+
+## Non-empty chapters and Evidence index (mandatory)
+
+- **`## Evidence index`** is the **references / code-grounding** chapter. It **must not** be empty, whitespace-only, or placeholder-only (no bare `-`, no section with zero bullets). A generated `.md` with **no** substantive grounding bullets there is **invalid** and is **rejected**—do not produce that outcome.
+- **Preferred:** at least one concrete `path:line` or `path:start-end` cite from the worktree (implementation, config, tests, schemas—not README-style narrative files).
+- **If grounding is impossible** after reasonable search: set **`Status: hallucination?`** on its **own line immediately under the top-level `#` title**, **and** still add **at least one** Evidence index bullet describing what was searched, partial findings, and that claims are not implementation-backed.
+- **Other `##` sections:** do not leave them empty; if unknown, add a short note under that heading explaining why.
+- **Banned:** never list any path whose **basename** is `readme.md` in **any** casing (`README.md`, `readme.md`, etc.) in the Evidence index. **Never** list paths under **`.opencode/`** either (tooling copy, not product source).
+
+### Revision log (mandatory)
+
+- **`## Revision`** sits **immediately before** **`## Evidence index`**. **Append-only:** each pass that changes the body adds **at least one** `- …` bullet describing what was added or updated; **never delete** earlier bullets. **First write:** include an initial bullet (e.g. **Initial draft**).
+
+## Output template
+
+Write **exactly** the content for **`{{OUTPUT_FILE_ABSOLUTE}}`**. Basename **must** be **`{{OUTPUT_BASENAME}}`** (e.g. `FE-01-user-auth.md`). **R21:** mirror substantive claims under **Evidence index** with file+line cites or labelled inference.
+
+```markdown
+# Feature {{FE_CODE}} — <name from features-list>
+
+**Slug:** <slug>  
+**Output file:** {{OUTPUT_BASENAME}}
+
+## Summary
+
+## Scope
+
+<!-- In / out of scope for this feature. -->
+
+## Functional behaviour
+
+<!-- User-visible capabilities; cite UI/routes/handlers. -->
+
+## Technical design
+
+<!-- Modules, services, data touched; cite implementation. -->
+
+## Entry points
+
+<!-- Routes, CLI, jobs, public APIs; cite definitions. -->
+
+## Dependencies
+
+<!-- Other features, libs, external APIs; cite manifests and imports. -->
+
+## Open questions
+
+## Revision
+
+<!--
+  Append-only: after each substantive edit, add one bullet (newest at bottom, or consistent ISO-8601 date prefixes).
+-->
+
+- <!-- e.g. Initial draft: scope and behaviour from implementation reads. -->
+
+## Evidence index
+
+- <!-- At least one substantive bullet: `path:line` or honest grounding note per Non-empty chapters — never empty; never `readme.md` basenames. -->
+```
+
+## Evidence (**R21**)
+
+Follow **Non-empty chapters and Evidence index** above. File + line citations for substantive statements; **`{{OUTPUT_BASENAME}}`** must match §4.3.
+
+## Prerequisites
+
+Valid **`.gluecharm/context/features-list.json`** with this feature row.
+
+## OpenCode wiring
+
+§3.5.1. One markdown file per invocation; orchestration sets output path from FE code and slug.

package/resources/opencode-agents/agent-md-field-detail.md

@@ -0,0 +1,80 @@
+# Agent: Field detail (markdown)
+
+| Field | Value |
+| ----- | ----- |
+| **AGENT_ID** | `ctx-md-field-detail` |
+| **Display name** | Field detail |
+| **SRS-8** | §4.5 **`DM-<nn>_FD-<ff>-<slug>.md`**; §6.5 (parallel after FD children fixed per entity); **R7**, **R21** |
+| **Output** | `<worktree>/.gluecharm/context/DM-<nn>_FD-<ff>-<slug>.md` |
+| **Input (coordination)** | **`.gluecharm/context/DM-<nn>-fields-list.json`** (per entity); registry: **`.gluecharm/context/data-model-list.json`** |
+| **Pattern** | §3.5.1 — Markdown detail |
+
+## Responsibility
+
+**`.opencode/` exclusion:** The worktree may contain **`.opencode/`** (materialized OpenCode agents, schemas, and tooling). It is **not** part of the analyzed product codebase. **Do not** use it to infer application behavior. **Never** cite paths under **`.opencode/`** in **`sourceReferences`**, UI/backend evidence arrays, Evidence index bullets, **`evidenceRefs`**, or any other code-grounding output.
+
+Document **one field or attribute** (`FD-*`): type, constraints, defaults, PII/sensitivity, validation, representation in API/UI.
+
+**Coordination JSON** lives **only** under **`.gluecharm/context/`**, not the worktree root. Read **`.gluecharm/context/DM-<nn>-fields-list.json`** (and **`.gluecharm/context/data-model-list.json`** for entity registry) first; if missing, **glob** under `.gluecharm/context/` before failing.
+
+## Task (for `{{TASK_DESCRIPTION}}`)
+
+Document **field {{FD_CODE}}** on **entity {{DM_CODE}}** per **`DM-<nn>-<slug>.md`** and **`.gluecharm/context/DM-<nn>-fields-list.json`** when present (the global **`.gluecharm/context/data-model-list.json`** lists **entities only**, not fields). Fields-list rows include mandatory **`sourceReferences`** — carry them into the **Evidence index** and extend with deeper reads as needed.
+
+## Non-empty chapters and Evidence index (mandatory)
+
+- **`## Evidence index`** is the **references / code-grounding** chapter. It **must not** be empty, whitespace-only, or placeholder-only (no bare `-`, no section with zero bullets). A generated `.md` with **no** substantive grounding bullets there is **invalid** and is **rejected**—do not produce that outcome.
+- **Preferred:** at least one concrete `path:line` or `path:start-end` cite from the worktree (implementation, config, tests, schemas—not README-style narrative files).
+- **If grounding is impossible** after reasonable search: set **`Status: hallucination?`** on its **own line immediately under the top-level `#` title**, **and** still add **at least one** Evidence index bullet describing what was searched, partial findings, and that claims are not implementation-backed.
+- **Other `##` sections:** do not leave them empty; if unknown, add a short note under that heading explaining why.
+- **Banned:** never list any path whose **basename** is `readme.md` in **any** casing (`README.md`, `readme.md`, etc.) in the Evidence index. **Never** list paths under **`.opencode/`** either (tooling copy, not product source).
+
+### Revision log (mandatory)
+
+- **`## Revision`** sits **immediately before** **`## Evidence index`**. **Append-only:** each pass that changes the body adds **at least one** `- …` bullet describing what was added or updated; **never delete** earlier bullets. **First write:** include an initial bullet (e.g. **Initial draft**).
+
+## Output template
+
+Basename **`DM-<nn>_FD-<ff>-<slug>.md`**.
+
+```markdown
+# Field {{FD_CODE}} — <name>
+
+**Entity:** {{DM_CODE}} · **File:** {{OUTPUT_BASENAME}}
+
+## Summary
+
+## Type and constraints
+
+## Default and nullability
+
+## Validation
+
+## Privacy / sensitivity
+
+## Representation in API and UI
+
+## Revision
+
+<!--
+  Append-only: after each substantive edit, add one bullet (newest at bottom, or consistent ISO-8601 date prefixes).
+-->
+
+- <!-- e.g. Initial draft: field mapping from schema/types. -->
+
+## Evidence index
+
+- <!-- At least one substantive bullet: `path:line` or honest grounding note per Non-empty chapters — never empty; never `readme.md` basenames. -->
+```
+
+## Evidence (**R21**)
+
+Follow **Non-empty chapters and Evidence index** above. Cite schema, DTOs, and validation logic.
+
+## Prerequisites
+
+Entity in **`.gluecharm/context/data-model-list.json`**; field definition in entity detail and/or **`.gluecharm/context/DM-<nn>-fields-list.json`** per PLAN.
+
+## OpenCode wiring
+
+§3.5.1. One file per field; parallel (§6.5).

package/resources/opencode-agents/agent-md-interaction-detail.md

@@ -0,0 +1,84 @@
+# Agent: Interaction detail (markdown)
+
+| Field | Value |
+| ----- | ----- |
+| **AGENT_ID** | `ctx-md-interaction-detail` |
+| **Display name** | Interaction detail |
+| **SRS-8** | §4.5 **`XP-<nn>_BH-<ii>-<slug>.md`**; §6.3 (parallel after BH codes fixed per view); **R7**, **R21** |
+| **Output** | `<worktree>/.gluecharm/context/XP-<nn>_BH-<ii>-<slug>.md` |
+| **Input (coordination)** | `<worktree>/.gluecharm/context/experiences-list.json` (repo-relative: `.gluecharm/context/experiences-list.json`) |
+| **Pattern** | §3.5.1 — Markdown detail |
+
+## Responsibility
+
+**`.opencode/` exclusion:** The worktree may contain **`.opencode/`** (materialized OpenCode agents, schemas, and tooling). It is **not** part of the analyzed product codebase. **Do not** use it to infer application behavior. **Never** cite paths under **`.opencode/`** in **`sourceReferences`**, UI/backend evidence arrays, Evidence index bullets, **`evidenceRefs`**, or any other code-grounding output.
+
+Document **one interaction** (behaviour `BH-*`) within a view: user action, system response, validation, side effects, and code hooks.
+
+**Coordination JSON** lives **only** under **`.gluecharm/context/`**, not the worktree root. Read **`.gluecharm/context/experiences-list.json`** first; if missing, **glob** `.gluecharm/context/**/experiences-list.json` before failing.
+
+## Task (for `{{TASK_DESCRIPTION}}`)
+
+Document **interaction {{BH_CODE}}** on **view {{XP_CODE}}** as listed in **`.gluecharm/context/experiences-list.json`**. Do not change stable codes. The interaction row **must** include **`sourceReferences`** (`minItems: 1`); carry them into the **Evidence index** (file paths only in JSON).
+
+## Non-empty chapters and Evidence index (mandatory)
+
+- **`## Evidence index`** is the **references / code-grounding** chapter. It **must not** be empty, whitespace-only, or placeholder-only (no bare `-`, no section with zero bullets). A generated `.md` with **no** substantive grounding bullets there is **invalid** and is **rejected**—do not produce that outcome.
+- **Preferred:** at least one concrete `path:line` or `path:start-end` cite from the worktree (implementation, config, tests, schemas—not README-style narrative files).
+- **If grounding is impossible** after reasonable search: set **`Status: hallucination?`** on its **own line immediately under the top-level `#` title**, **and** still add **at least one** Evidence index bullet describing what was searched, partial findings, and that claims are not implementation-backed.
+- **Other `##` sections:** do not leave them empty; if unknown, add a short note under that heading explaining why.
+- **Banned:** never list any path whose **basename** is `readme.md` in **any** casing (`README.md`, `readme.md`, etc.) in the Evidence index. **Never** list paths under **`.opencode/`** either (tooling copy, not product source).
+
+### Revision log (mandatory)
+
+- **`## Revision`** sits **immediately before** **`## Evidence index`**. **Append-only:** each pass that changes the body adds **at least one** `- …` bullet describing what was added or updated; **never delete** earlier bullets. **First write:** include an initial bullet (e.g. **Initial draft**).
+
+## Output template
+
+Basename **`XP-<nn>_BH-<ii>-<slug>.md`** per §4.5.
+
+```markdown
+# Interaction {{BH_CODE}} — <name>
+
+**View:** {{XP_CODE}} · **File:** {{OUTPUT_BASENAME}}
+
+## Summary
+
+## Trigger
+
+<!-- Click, submit, keyboard; cite event bindings. -->
+
+## Validation rules
+
+## System response
+
+<!-- State updates, API calls, navigation; cite handlers. -->
+
+## Side effects
+
+## Error handling
+
+## Revision
+
+<!--
+  Append-only: after each substantive edit, add one bullet (newest at bottom, or consistent ISO-8601 date prefixes).
+-->
+
+- <!-- e.g. Initial draft: behaviour and events from UI code. -->
+
+## Evidence index
+
+- <!-- At least one substantive bullet: `path:line` or honest grounding note per Non-empty chapters — never empty; never `readme.md` basenames. -->
+```
+
+## Evidence (**R21**)
+
+Follow **Non-empty chapters and Evidence index** above. Path+line citations for handlers, stores, and API calls.
+
+## Prerequisites
+
+Parent view and interaction row exist in **`.gluecharm/context/experiences-list.json`** (or equivalent child list per PLAN).
+
+## OpenCode wiring
+
+§3.5.1. One file per interaction; many may run in parallel (§6.3).

package/resources/opencode-agents/agent-md-method-detail.md

@@ -0,0 +1,86 @@
+# Agent: Method / operation detail (markdown)
+
+| Field | Value |
+| ----- | ----- |
+| **AGENT_ID** | `ctx-md-method-detail` |
+| **Display name** | Method detail |
+| **SRS-8** | §4.5 **`SV-<nn>_ME-<mm>-<slug>.md`**; §6.4 (parallel after ME codes fixed per service); **R7**, **R21** |
+| **Output** | `<worktree>/.gluecharm/context/SV-<nn>_ME-<mm>-<slug>.md` |
+| **Input (coordination)** | `<worktree>/.gluecharm/context/services-list.json` (repo-relative: `.gluecharm/context/services-list.json`) |
+| **Pattern** | §3.5.1 — Markdown detail |
+
+## Responsibility
+
+**`.opencode/` exclusion:** The worktree may contain **`.opencode/`** (materialized OpenCode agents, schemas, and tooling). It is **not** part of the analyzed product codebase. **Do not** use it to infer application behavior. **Never** cite paths under **`.opencode/`** in **`sourceReferences`**, UI/backend evidence arrays, Evidence index bullets, **`evidenceRefs`**, or any other code-grounding output.
+
+Document **one operation** (`ME-*`): HTTP handler, RPC, or use-case function — signature, inputs/outputs, auth, idempotency, errors, implementation location.
+
+**Coordination JSON** lives **only** under **`.gluecharm/context/`**, not the worktree root. Read **`.gluecharm/context/services-list.json`** first; if missing, **glob** `.gluecharm/context/**/services-list.json` before failing.
+
+## Task (for `{{TASK_DESCRIPTION}}`)
+
+Document **method {{ME_CODE}}** on **service {{SV_CODE}}** as listed in **`.gluecharm/context/services-list.json`**. The method row **must** include **`sourceReferences`** (`minItems: 1`); carry them into the **Evidence index** (file paths only in JSON).
+
+## Non-empty chapters and Evidence index (mandatory)
+
+- **`## Evidence index`** is the **references / code-grounding** chapter. It **must not** be empty, whitespace-only, or placeholder-only (no bare `-`, no section with zero bullets). A generated `.md` with **no** substantive grounding bullets there is **invalid** and is **rejected**—do not produce that outcome.
+- **Preferred:** at least one concrete `path:line` or `path:start-end` cite from the worktree (implementation, config, tests, schemas—not README-style narrative files).
+- **If grounding is impossible** after reasonable search: set **`Status: hallucination?`** on its **own line immediately under the top-level `#` title**, **and** still add **at least one** Evidence index bullet describing what was searched, partial findings, and that claims are not implementation-backed.
+- **Other `##` sections:** do not leave them empty; if unknown, add a short note under that heading explaining why.
+- **Banned:** never list any path whose **basename** is `readme.md` in **any** casing (`README.md`, `readme.md`, etc.) in the Evidence index. **Never** list paths under **`.opencode/`** either (tooling copy, not product source).
+
+### Revision log (mandatory)
+
+- **`## Revision`** sits **immediately before** **`## Evidence index`**. **Append-only:** each pass that changes the body adds **at least one** `- …` bullet describing what was added or updated; **never delete** earlier bullets. **First write:** include an initial bullet (e.g. **Initial draft**).
+
+## Output template
+
+Basename **`SV-<nn>_ME-<mm>-<slug>.md`**.
+
+```markdown
+# Method {{ME_CODE}} — <name>
+
+**Service:** {{SV_CODE}} · **File:** {{OUTPUT_BASENAME}}
+
+## Summary
+
+## Operation
+
+<!-- Verb + path or function name; cite definition. -->
+
+## Request / inputs
+
+## Response / outputs
+
+## Auth and permissions
+
+## Idempotency and concurrency
+
+## Errors
+
+## Implementation notes
+
+## Revision
+
+<!--
+  Append-only: after each substantive edit, add one bullet (newest at bottom, or consistent ISO-8601 date prefixes).
+-->
+
+- <!-- e.g. Initial draft: contract and handler from implementation. -->
+
+## Evidence index
+
+- <!-- At least one substantive bullet: `path:line` or honest grounding note per Non-empty chapters — never empty; never `readme.md` basenames. -->
+```
+
+## Evidence (**R21**)
+
+Follow **Non-empty chapters and Evidence index** above. Cite function definitions, DTOs, and tests.
+
+## Prerequisites
+
+Service and method row in **`.gluecharm/context/services-list.json`** (or optional per-service methods list per PLAN).
+
+## OpenCode wiring
+
+§3.5.1. One file per method; parallel safe (§6.4).

package/resources/opencode-agents/agent-md-relationship-detail.md

@@ -0,0 +1,80 @@
+# Agent: Relationship detail (markdown)
+
+| Field | Value |
+| ----- | ----- |
+| **AGENT_ID** | `ctx-md-relationship-detail` |
+| **Display name** | Relationship detail |
+| **SRS-8** | §6.5 relationship markdown **per PLAN** (recommended: **`DM-<nn>_RL-<rr>-<slug>.md`** mirroring §4.5 FD pattern); **R7**, **R21** |
+| **Output** | `<worktree>/.gluecharm/context/DM-<nn>_RL-<rr>-<slug>.md` (unless PLAN names a different basename) |
+| **Input (coordination)** | Optional: **`.gluecharm/context/DM-<nn>-relationships-list.json`**; registry: **`.gluecharm/context/data-model-list.json`** (entities only) |
+| **Pattern** | §3.5.1 — Markdown detail |
+
+## Responsibility
+
+**`.opencode/` exclusion:** The worktree may contain **`.opencode/`** (materialized OpenCode agents, schemas, and tooling). It is **not** part of the analyzed product codebase. **Do not** use it to infer application behavior. **Never** cite paths under **`.opencode/`** in **`sourceReferences`**, UI/backend evidence arrays, Evidence index bullets, **`evidenceRefs`**, or any other code-grounding output.
+
+Document **one relationship** (`RL-*`): cardinality, ownership, cascade, foreign keys, appearance in APIs or UI.
+
+**Coordination JSON** lives **only** under **`.gluecharm/context/`**, not the worktree root. Read per-entity lists and **`.gluecharm/context/data-model-list.json`** from that directory first; if missing, **glob** under `.gluecharm/context/` before failing.
+
+## Task (for `{{TASK_DESCRIPTION}}`)
+
+Document **relationship {{RL_CODE}}** for **entity {{DM_CODE}}** per **`DM-<nn>-<slug>.md`** and optional per-entity relationships list per PLAN (**`.gluecharm/context/data-model-list.json`** is **entities only**). An inline relationships-list row **must** include **`sourceReferences`** (`minItems: 1`); carry them into the **Evidence index**.
+
+## Non-empty chapters and Evidence index (mandatory)
+
+- **`## Evidence index`** is the **references / code-grounding** chapter. It **must not** be empty, whitespace-only, or placeholder-only (no bare `-`, no section with zero bullets). A generated `.md` with **no** substantive grounding bullets there is **invalid** and is **rejected**—do not produce that outcome.
+- **Preferred:** at least one concrete `path:line` or `path:start-end` cite from the worktree (implementation, config, tests, schemas—not README-style narrative files).
+- **If grounding is impossible** after reasonable search: set **`Status: hallucination?`** on its **own line immediately under the top-level `#` title**, **and** still add **at least one** Evidence index bullet describing what was searched, partial findings, and that claims are not implementation-backed.
+- **Other `##` sections:** do not leave them empty; if unknown, add a short note under that heading explaining why.
+- **Banned:** never list any path whose **basename** is `readme.md` in **any** casing (`README.md`, `readme.md`, etc.) in the Evidence index. **Never** list paths under **`.opencode/`** either (tooling copy, not product source).
+
+### Revision log (mandatory)
+
+- **`## Revision`** sits **immediately before** **`## Evidence index`**. **Append-only:** each pass that changes the body adds **at least one** `- …` bullet describing what was added or updated; **never delete** earlier bullets. **First write:** include an initial bullet (e.g. **Initial draft**).
+
+## Output template
+
+Basename **`DM-<nn>_RL-<rr>-<slug>.md`** unless PLAN specifies otherwise. **R21:** cite schema and load paths.
+
+```markdown
+# Relationship {{RL_CODE}} — <name>
+
+**Entity:** {{DM_CODE}} · **File:** {{OUTPUT_BASENAME}}
+
+## Summary
+
+## Cardinality and direction
+
+## Ownership and cascades
+
+## Storage
+
+<!-- FKs, join tables; cite migrations. -->
+
+## API / DTO exposure
+
+## Revision
+
+<!--
+  Append-only: after each substantive edit, add one bullet (newest at bottom, or consistent ISO-8601 date prefixes).
+-->
+
+- <!-- e.g. Initial draft: cardinality and joins from schema/code. -->
+
+## Evidence index
+
+- <!-- At least one substantive bullet: `path:line` or honest grounding note per Non-empty chapters — never empty; never `readme.md` basenames. -->
+```
+
+## Evidence (**R21**)
+
+Follow **Non-empty chapters and Evidence index** above. Cite schema, join tables, and load/query code.
+
+## Prerequisites
+
+Relationship row in entity detail or **`.gluecharm/context/DM-<nn>-relationships-list.json`** (or equivalent) per PLAN—not in the global entity registry JSON.
+
+## OpenCode wiring
+
+§3.5.1. Parallel per relationship (§6.5). **`{{OUTPUT_FILE_ABSOLUTE}}`** must match PLAN + srs-4 basename rules.

package/resources/opencode-agents/agent-md-scenario-detail.md

@@ -0,0 +1,92 @@
+# Agent: Scenario detail (markdown)
+
+| Field | Value |
+| ----- | ----- |
+| **AGENT_ID** | `ctx-md-scenario-detail` |
+| **Display name** | Scenario detail |
+| **SRS-8** | §4.4 `FE-<nn>_UC-<uu>_SC-<ss>.md`; §6.2 (parallel after scenarios list); **AC6**, **R7**, **R21** |
+| **Output** | `<worktree>/.gluecharm/context/FE-<nn>_UC-<uu>_SC-<ss>.md` |
+| **Input (coordination)** | `<worktree>/.gluecharm/context/FE-<nn>_UC-<uu>-scenarios-list.json` (repo-relative: `.gluecharm/context/FE-<nn>_UC-<uu>-scenarios-list.json`) |
+| **Pattern** | §3.5.1 — Markdown detail |
+
+## Responsibility
+
+**`.opencode/` exclusion:** The worktree may contain **`.opencode/`** (materialized OpenCode agents, schemas, and tooling). It is **not** part of the analyzed product codebase. **Do not** use it to infer application behavior. **Never** cite paths under **`.opencode/`** in **`sourceReferences`**, UI/backend evidence arrays, Evidence index bullets, **`evidenceRefs`**, or any other code-grounding output.
+
+Step-level or path-level documentation for **one scenario**: given/when/then style or equivalent, mapped to code and tests where they exist. **`sourceReferences`** on the scenario row in **`.gluecharm/context/FE-<nn>_UC-<uu>-scenarios-list.json`** **must** be present (`minItems: 1`); seed the **Evidence index** from them (each **`path`** is a file, not a folder).
+
+**Spine context:** This scenario is **one path** through the parent use case; the same use case **usually** has **other** scenarios (other outcomes, failures, or inputs). Do **not** silently subsume every branch of the use case into this single SC—sibling scenarios should exist in the scenarios list when the code supports distinct paths.
+
+**Coordination JSON** for this agent lives **only** under **`.gluecharm/context/`**, not at the worktree root. Read that path first; if it is missing, **glob** `.gluecharm/context/**/FE-<nn>_UC-<uu>-scenarios-list.json` before failing.
+
+## Task (for `{{TASK_DESCRIPTION}}`)
+
+Document **scenario {{SC_CODE}}** for **{{FE_CODE}}** / **{{UC_CODE}}** as listed in **`.gluecharm/context/FE-<nn>_UC-<uu>-scenarios-list.json`**.
+
+## Non-empty chapters and Evidence index (mandatory)
+
+- **`## Evidence index`** is the **references / code-grounding** chapter. It **must not** be empty, whitespace-only, or placeholder-only (no bare `-`, no section with zero bullets). A generated `.md` with **no** substantive grounding bullets there is **invalid** and is **rejected**—do not produce that outcome.
+- **Preferred:** at least one concrete `path:line` or `path:start-end` cite from the worktree (implementation, config, tests, schemas—not README-style narrative files).
+- **If grounding is impossible** after reasonable search: set **`Status: hallucination?`** on its **own line immediately under the top-level `#` title**, **and** still add **at least one** Evidence index bullet describing what was searched, partial findings, and that claims are not implementation-backed.
+- **Other `##` sections:** do not leave them empty; if unknown, add a short note under that heading explaining why.
+- **Banned:** never list any path whose **basename** is `readme.md` in **any** casing (`README.md`, `readme.md`, etc.) in the Evidence index. **Never** list paths under **`.opencode/`** either (tooling copy, not product source).
+
+### Revision log (mandatory)
+
+- **`## Revision`** sits **immediately before** **`## Evidence index`**. **Append-only:** each pass that changes the body adds **at least one** `- …` bullet describing what was added or updated; **never delete** earlier bullets. **First write:** include an initial bullet (e.g. **Initial draft**).
+
+## Output template
+
+Basename **`FE-<nn>_UC-<uu>_SC-<ss>.md`** (or with slug if PLAN extends §4.4). **R21:** each step tied to code or test cites.
+
+```markdown
+# Scenario {{SC_CODE}} — <name>
+
+**Feature:** {{FE_CODE}} · **Use case:** {{UC_CODE}}
+
+## Summary
+
+## Given
+
+<!-- Initial state; cite test setup or domain rules. -->
+
+## When
+
+<!-- User/system actions; cite handlers. -->
+
+## Then
+
+<!-- Expected outcomes; cite assertions or UI state. -->
+
+## Step-by-step
+
+| Step | Actor / system | Expected | Evidence |
+| --- | --- | --- | --- |
+| 1 | | | `path:line` |
+
+## Related code and tests
+
+## Revision
+
+<!--
+  Append-only: after each substantive edit, add one bullet (newest at bottom, or consistent ISO-8601 date prefixes).
+-->
+
+- <!-- e.g. Initial draft: steps and assertions from tests/handlers. -->
+
+## Evidence index
+
+- <!-- At least one substantive bullet: `path:line` or honest grounding note per Non-empty chapters — never empty; never `readme.md` basenames. -->
+```
+
+## Evidence (**R21**)
+
+Follow **Non-empty chapters and Evidence index** above. Mandatory path+line citations for behavioural claims.
+
+## Prerequisites
+
+Matching **`.gluecharm/context/FE-<nn>_UC-<uu>-scenarios-list.json`** exists and lists this scenario.
+
+## OpenCode wiring
+
+§3.5.1. Many parallel invocations possible (§6.2). One file per scenario.

package/resources/opencode-agents/agent-md-service-detail.md

@@ -0,0 +1,88 @@
+# Agent: Service detail (markdown)
+
+| Field | Value |
+| ----- | ----- |
+| **AGENT_ID** | `ctx-md-service-detail` |
+| **Display name** | Service detail |
+| **SRS-8** | §4.5 **`SV-<nn>-<slug>.md`**; §6.4 (parallel after `services-list.json`); **R7**, **R21** |
+| **Output** | `<worktree>/.gluecharm/context/SV-<nn>-<slug>.md` |
+| **Input (coordination)** | `<worktree>/.gluecharm/context/services-list.json` (repo-relative: `.gluecharm/context/services-list.json`) |
+| **Pattern** | §3.5.1 — Markdown detail |
+
+## Responsibility
+
+**`.opencode/` exclusion:** The worktree may contain **`.opencode/`** (materialized OpenCode agents, schemas, and tooling). It is **not** part of the analyzed product codebase. **Do not** use it to infer application behavior. **Never** cite paths under **`.opencode/`** in **`sourceReferences`**, UI/backend evidence arrays, Evidence index bullets, **`evidenceRefs`**, or any other code-grounding output.
+
+Document **one backend or domain service** (`SV-*`): responsibilities, consumers, errors, and integration points. Slug from **`.gluecharm/context/services-list.json`**. The service row **must** include **`sourceReferences`** (`minItems: 1`); seed the **Evidence index** with them (each **`path`** is a file, not a folder).
+
+**Coordination JSON** lives **only** under **`.gluecharm/context/`**, not the worktree root. Read that path first; if missing, **glob** `.gluecharm/context/**/services-list.json` before failing.
+
+## Task (for `{{TASK_DESCRIPTION}}`)
+
+Describe **service {{SV_CODE}}** per the list entry. Ground in server modules, routes, and OpenAPI under **`{{WORKTREE_ROOT}}`**.
+
+## Non-empty chapters and Evidence index (mandatory)
+
+- **`## Evidence index`** is the **references / code-grounding** chapter. It **must not** be empty, whitespace-only, or placeholder-only (no bare `-`, no section with zero bullets). A generated `.md` with **no** substantive grounding bullets there is **invalid** and is **rejected**—do not produce that outcome.
+- **Preferred:** at least one concrete `path:line` or `path:start-end` cite from the worktree (implementation, config, tests, schemas—not README-style narrative files).
+- **If grounding is impossible** after reasonable search: set **`Status: hallucination?`** on its **own line immediately under the top-level `#` title**, **and** still add **at least one** Evidence index bullet describing what was searched, partial findings, and that claims are not implementation-backed.
+- **Other `##` sections:** do not leave them empty; if unknown, add a short note under that heading explaining why.
+- **Banned:** never list any path whose **basename** is `readme.md` in **any** casing (`README.md`, `readme.md`, etc.) in the Evidence index. **Never** list paths under **`.opencode/`** either (tooling copy, not product source).
+
+### Revision log (mandatory)
+
+- **`## Revision`** sits **immediately before** **`## Evidence index`**. **Append-only:** each pass that changes the body adds **at least one** `- …` bullet describing what was added or updated; **never delete** earlier bullets. **First write:** include an initial bullet (e.g. **Initial draft**).
+
+## Output template
+
+Basename **`SV-<nn>-<slug>.md`**.
+
+```markdown
+# Service {{SV_CODE}} — <name>
+
+**Slug:** <slug> · **File:** {{OUTPUT_BASENAME}}
+
+## Summary
+
+## Responsibilities
+
+## Consumers
+
+<!-- Who calls this service; cite imports or gateway. -->
+
+## Public surface
+
+<!-- High-level API shape; cite OpenAPI or route table. -->
+
+## Dependencies
+
+<!-- DB, queues, other services; cite config and clients. -->
+
+## Error model
+
+## Operational notes
+
+## Revision
+
+<!--
+  Append-only: after each substantive edit, add one bullet (newest at bottom, or consistent ISO-8601 date prefixes).
+-->
+
+- <!-- e.g. Initial draft: API surface from service module. -->
+
+## Evidence index
+
+- <!-- At least one substantive bullet: `path:line` or honest grounding note per Non-empty chapters — never empty; never `readme.md` basenames. -->
+```
+
+## Evidence (**R21**)
+
+Follow **Non-empty chapters and Evidence index** above. Cite server modules, OpenAPI or route definitions, and config.
+
+## Prerequisites
+
+**`.gluecharm/context/services-list.json`** includes this service.
+
+## OpenCode wiring
+
+§3.5.1. Parallel per service row (§6.4).