Sync ia skill with upstream (v1.1.0)

.agents/plugins/marketplace.json

@@ -8,7 +8,7 @@
   "plugins": [
     {
       "name": "ia",
-      "version": "1.0.2",
+      "version": "1.1.0",
       "description": "IBM i Impact Analysis — dependency tracing, field impact, call hierarchy, program documentation.",
       "source": "./"
     }

.claude-plugin/marketplace.json

@@ -8,7 +8,7 @@
   "plugins": [
     {
       "name": "ia",
-      "version": "1.0.2",
+      "version": "1.1.0",
       "description": "IBM i Impact Analysis — dependency tracing, field impact, call hierarchy, program documentation.",
       "source": "./"
     }

.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "ia",
-  "version": "1.0.2",
+  "version": "1.1.0",
   "description": "IBM i Impact Analysis — dependency tracing, field impact, call hierarchy, program documentation.",
   "homepage": "https://github.com/programmersio-ibmi/skills",
   "license": "MIT",

.codex-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "ia",
-  "version": "1.0.2",
+  "version": "1.1.0",
   "description": "IBM i Impact Analysis for coding agents.",
   "homepage": "https://github.com/programmersio-ibmi/skills",
   "license": "MIT",

.cursor-plugin/plugin.json

@@ -1,7 +1,7 @@
 {
   "name": "ia",
   "displayName": "iA — IBM i Impact Analysis",
-  "version": "1.0.2",
+  "version": "1.1.0",
   "description": "IBM i Impact Analysis — dependency tracing, field impact, call hierarchy, program documentation.",
   "publisher": "programmersio",
   "homepage": "https://github.com/programmersio-ibmi/skills",

.github/plugin/marketplace.json

@@ -8,7 +8,7 @@
   "plugins": [
     {
       "name": "ia",
-      "version": "1.0.2",
+      "version": "1.1.0",
       "description": "IBM i Impact Analysis — dependency tracing, field impact, call hierarchy, program documentation.",
       "source": "./"
     }

CHANGELOG.md

@@ -5,6 +5,18 @@ All notable changes to this project are documented here.
 The format is loosely based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
 and the project uses [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [1.1.0] — 2026-06-11
+
+### Added
+
+- **Visual flowchart generation** (`references/flowchart.md`): produce a single-page HTML flowchart of a program — caller context, subroutine/procedure flow, and external-call nodes — from `references/templates/flowchart-template.html`, gated by the `scripts/validate_flowchart.py` lint check.
+- **Test case document generation** (`references/test-case-generation.md`): a fifth document type (QA/UAT test scripts) alongside the four audience templates, built from `references/templates/template-testcases.md` and gated by the `scripts/validate_testcases.py` lint check.
+
+### Changed
+
+- Richer Word/PDF export from the markdown converter scripts: branded cover page, styled tables/headings, and broader markdown coverage in both `convert_md_to_docx.py` and `convert_md_to_pdf.py`.
+- `SKILL.md` references table, templates README, and the program-documentation workflow refreshed to route to the new flowchart and test-case capabilities.
+
 ## [1.0.2] — 2026-06-02
 
 ### Changed

gemini-extension.json

@@ -1,6 +1,6 @@
 {
   "name": "ia",
-  "version": "1.0.2",
+  "version": "1.1.0",
   "description": "IBM i Impact Analysis — dependency tracing, field impact, call hierarchy, program documentation.",
   "contextFileName": "GEMINI.md"
 }

package.json

@@ -1,6 +1,6 @@
 {
   "name": "ia-tools-skill",
-  "version": "1.0.2",
+  "version": "1.1.0",
   "description": "IBM i Impact Analysis — dependency tracing, field impact, call hierarchy, program documentation.",
   "license": "MIT",
   "type": "module",

plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "ia",
-  "version": "1.0.2",
+  "version": "1.1.0",
   "description": "IBM i Impact Analysis — dependency tracing, field impact, call hierarchy, program documentation.",
   "publisher": "programmersio",
   "homepage": "https://github.com/programmersio-ibmi/skills",

skills/ia/SKILL.md

@@ -141,3 +141,5 @@ Do not attempt to diagnose server-side issues or retry indefinitely.
 | Complex analysis chains | [query-flows.md](references/query-flows.md) |
 | Analysis playbooks | [playbook.md](references/playbook.md) |
 | Program documentation | [program-documentation.md](references/program-documentation.md) |
+| Test case document for a program (QA/UAT scripts) | [test-case-generation.md](references/test-case-generation.md) |
+| Visual flowchart of a program (single-page HTML) | [flowchart.md](references/flowchart.md) |

skills/ia/references/flowchart.md

@@ -0,0 +1,146 @@
+# Flowchart Generation
+
+Use this when a user asks for a **visual flowchart** of an IBM i program ("flowchart for X", "diagram the logic of X", "visualize the flow"). The deliverable is **one self-contained HTML file** holding a **single top-to-bottom Mermaid diagram** of the whole program, plus a subroutine index and file list.
+
+> **One diagram, HTML only.** Mermaid renders live in the browser, so it stays out of the text-based Word/PDF spec pipeline (which uses ASCII trees — see [program-documentation.md](program-documentation.md), Step 6). For a full written spec, use program documentation; for a visual flow, use this.
+
+---
+
+## 1. Gather the data (reuse the spec tools)
+
+One call gets almost everything:
+
+```
+ia_program_spec_bundle(program_name=X)                  → LOOKUP, COMPLEXITY, FILES, SUBROUTINES, CALLEES, PARAMS
+ia_subroutines(member_name=X)                           → BEGSR with exact line numbers (for the "L<n>" refs)
+ia_call_hierarchy(program_name=X, direction='CALLERS')  → who calls X (the "Called by" meta line)
+```
+
+If LOOKUP returns multiple versions, **stop and ask which library** (same rule as program docs). Read the **source** (`ia_rpg_source` / `ia_cl_source`) to follow the actual control flow — the diagram must reflect what the code does, not a guess. Collect: program name/library/type, total + executable lines, last-changed date, the subroutine list with line numbers, DB files, external calls, callers, and the entry interface (`*ENTRY`/PI parameters from PARAMS — or "no parameters — interactive"). While reading the source, also note the **F-keys** the program handles (F3 exit, F4 prompt, F12 cancel, …) for the F-key row.
+
+> **Procedure-based programs** (COMPLEXITY shows `PROC>0` with `SBR=0` — common in SQLRPGLE): `ia_subroutines` and the bundle's `SUBROUTINES` section come back empty. Build the blocks from the program's **procedures** instead — list them via `ia_procedure_params` (`PROCEDURE_NAME`) / `ia_procedure_xref`, and read source (`dcl-proc`) for line ranges.
+
+---
+
+## 2. Design ONE top-to-bottom diagram
+
+A single `flowchart TD`, the **whole program in one graph**, reading top → bottom in execution order: `START` at the top, the mainline/menu, then each branch flowing **downward**, and `END` reached from the program's exit.
+
+- **Stay at a high altitude — ~15–25 nodes for the whole program.** Show the mainline and each option/branch's key steps (the screens, the decisions, the DB I/O). **Collapse helper subroutines** (clear/reset/refresh, window load+display) into one representative node — the *full* subroutine list lives in the table below the diagram, not in boxes. This is what lets one diagram cover even a large program.
+- **No `subgraph`s.** They make Mermaid cluster areas side-by-side and destroy the top-to-bottom flow. Keep it one flat graph; the `class` colors already group nodes visually.
+- **External program calls are never collapsed.** Every `CALL`/`CALLP` to another *program* gets its own node in subroutine shape `[["CALL PGMX<br/>L.."]]` with class `extcall` — the hand-off to another program is exactly what a new developer must see.
+- **Loops read downward:** draw loop-backs (e.g. "return to menu after each action") as **dotted** edges `-.->|loop|` to the loop entry, and the program exit as a **thick** edge `==>|Yes|` to `END`. The solid forward arrows then carry the eye straight down.
+- **One `START`, one `END`.** Every block carries a line reference (`<br/>L131` or `<br/>L272-320`).
+- Decisions are diamonds; label both outgoing edges.
+
+---
+
+## 3. Mermaid rules (v10.9.0)
+
+The template pins `mermaid@10.9.0` — **do not change it** (v11.x has SVG layout issues).
+
+```
+flowchart TD
+    NODE(["Terminal"])        ← START / END (rounded)
+    NODE["Process"]           ← subroutine / process / message (rectangle)
+    NODE[/"Input/Output"/]    ← EXFMT / DISPLAY screen (parallelogram)
+    NODE[("Database")]        ← CHAIN / READ / WRITE / SETLL (cylinder)
+    NODE[["CALL PGMX"]]       ← external program call (subroutine shape)
+    NODE{"Decision?"}         ← IF / loop test / validation (diamond)
+    A --> B                   ← forward step
+    A -->|Label| B            ← labelled step
+    A -.->|loop| B            ← loop-back (dotted)
+    A ==>|Yes| B              ← program exit (thick)
+```
+
+- Always `flowchart TD`, never `graph TD`. Double quotes in labels; line breaks with `<br/>`.
+- Node names: alphanumeric + underscore only. Avoid the reserved word `end` as a node id (use `ENDP`).
+- Escape `&` as `&amp;`. Keep labels short — push detail into the table.
+
+---
+
+## 4. Colours — define once with `classDef`
+
+Put these **seven `classDef` lines once** at the end of the diagram, then assign nodes with `class`. (Don't repeat per-node `style` lines.) Dark fills + white text are WCAG-AA.
+
+```
+classDef terminal fill:#2E7D32,stroke:#1B5E20,color:#fff,stroke-width:2px;
+classDef process  fill:#1565C0,stroke:#0D47A1,color:#fff,stroke-width:2px;
+classDef display  fill:#0277BD,stroke:#01579B,color:#fff,stroke-width:2px;
+classDef db       fill:#6A1B9A,stroke:#4A148C,color:#fff,stroke-width:2px;
+classDef extcall  fill:#00695C,stroke:#004D40,color:#fff,stroke-width:2px;
+classDef decision fill:#F57F17,stroke:#E65100,color:#fff,stroke-width:2px;
+classDef error    fill:#C2185B,stroke:#880E4F,color:#fff,stroke-width:2px;
+class START,ENDP terminal;
+class A,B process;
+...
+```
+
+| Class | Use for |
+|-------|---------|
+| `terminal` | START / END / ENTRY / RETURN |
+| `process` | subroutine / process / status message |
+| `display` | EXFMT / screen / subfile |
+| `db` | CHAIN / READ / WRITE / SETLL / SQL |
+| `extcall` | CALL / CALLP to another program |
+| `decision` | IF / loop test / validation |
+| `error` | error path / failure message |
+
+---
+
+## 5. Build the page (copy the template)
+
+**Copy the template — never write the HTML from scratch:**
+
+```
+templates/flowchart-template.html  →  docs/program-specs/{PROGRAM_NAME}/{PROGRAM_NAME}_Flowchart.html
+```
+
+Replace only the content; leave the CSS and the one-line `mermaid.initialize` script alone. The page has exactly these parts:
+
+1. **Title / `<h1>` / one-line subtitle.**
+2. **Meta strip** — line 1: library · source member · type · lines (total + exec) · subroutine count · files · last updated. **Line 2:** Called by `<callers or none>` · Calls `<callees or none>` · Entry: `<*ENTRY/PI params, or "no parameters — interactive">`.
+3. **The diagram** (§2–4).
+4. **Legend** — the seven colours + a note that dotted = loop-back, **plus an F-key row** listing only the keys the program actually handles (`F3 Exit · F4 Lookup window · F12 Back to menu`).
+5. **What it does** — 2–3 sentences in plain language.
+6. **Subroutines (source index)** — one row per BEGSR (or procedure): name · line · area · **in diagram** (the node that represents it — its own node, or the node it's collapsed under) · purpose. *This is where the full detail lives.*
+7. **Files** — file · type/access · record format(s) · role.
+8. **Footer** — print note + "flowchart by iA from programmers.io".
+
+There are **no tabs, no zoom controls, and no print hook** — a single diagram renders on load, so `Ctrl+P` prints/saves the whole page (the print CSS just scales the SVG to the page width).
+
+---
+
+## 6. Verify before delivery — two gates
+
+**Gate 1 — lint (mandatory, run first):**
+
+```
+python scripts/validate_flowchart.py docs/program-specs/{PGM}/{PGM}_Flowchart.html
+```
+
+Must exit 0. It catches: not `flowchart TD`, `subgraph`s, the reserved `end` node id, leftover template placeholders, nodes with no `class` assigned, an unknown/undefined class, an unpinned Mermaid version, and bare `&` in labels.
+
+**Gate 2 — browser:** open the file and confirm:
+
+- [ ] The diagram **flows top → bottom**: `START` at top, branches descending, `END` reached via the exit edge; dotted arrows are the loop-backs.
+- [ ] It renders (not raw code); colours show.
+- [ ] Altitude is right — roughly 15–25 nodes; helper routines are collapsed, not drawn individually.
+- [ ] Every external `CALL` appears as a teal `[["CALL …"]]` node — program calls are never collapsed.
+- [ ] **Every** subroutine/procedure appears in the index table with its line number, and its **In diagram** column names a real node from the diagram.
+- [ ] Meta strip (both lines), F-key row, and file list match the iA data and the source.
+- [ ] **Ctrl+P** preview shows the whole diagram on the page.
+
+Save to `docs/program-specs/{PROGRAM_NAME}/{PROGRAM_NAME}_Flowchart.html`.
+
+## Common errors
+
+| Symptom | Fix |
+|---------|-----|
+| Areas sit side-by-side, flow isn't vertical | Remove `subgraph`s — use one flat `flowchart TD`. |
+| Diagram is a wall of boxes | Too low-altitude: collapse helper subroutines into representative nodes; move detail to the table. |
+| Mermaid syntax error | `flowchart TD` (not `graph TD`), double quotes, `<br/>` for breaks, no node id named `end`, `classDef` lines last. |
+| Text invisible on a node | You used a bare node with no `class` — assign one of the seven classes. |
+| An external CALL is invisible in the flow | Give it its own `[["CALL PGMX"]]` node with class `extcall` — never fold a program call into another node. |
+| Lint (Gate 1) fails | Read its findings — each maps to a row above or a leftover template placeholder. Fix and re-run before the browser check. |
+| Diagram broke after edits | You changed the `<script>` or mermaid version — revert to the template's. |