add orchestrated agent skills and workflow by tdejoigny-ledger · Pull Request #195 · LedgerHQ/app-boilerplate

tdejoigny-ledger · 2026-02-17T17:18:18Z

No description provided.

codecov · 2026-02-17T17:19:53Z

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 90.81%. Comparing base (ac10944) to head (4927426).
⚠️ Report is 29 commits behind head on master.

Additional details and impacted files

@@           Coverage Diff           @@
##           master     #195   +/-   ##
=======================================
  Coverage   90.81%   90.81%           
=======================================
  Files           5        5           
  Lines          98       98           
  Branches       22       22           
=======================================
  Hits           89       89           
  Misses          9        9

Flag	Coverage Δ
unittests	`90.81% <ø> (ø)`

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

Copilot

Pull request overview

This PR introduces a multi-agent prompt framework under .github/prompts/ to orchestrate Ledger app development across implementation (C), QA (Ragger/Speculos), documentation, and review “gatekeeping”.

Changes:

Added an Orchestrator prompt defining a 4-phase Dev → QA → Doc → Review pipeline.
Added four “skill” prompt files specifying role responsibilities and checklists for C, QA, Doc, and Review agents.
Added a .github/prompts/README.md describing how to use the prompt suite.

Reviewed changes

Copilot reviewed 6 out of 6 changed files in this pull request and generated 12 comments.

Show a summary per file

File	Description
.github/prompts/ORCHESTRATOR.md	Defines the multi-phase orchestration workflow and logging template.
.github/prompts/SKILL_LEDGER_C.md	Dev-agent rules for embedded C implementation/build expectations.
.github/prompts/SKILL_LEDGER_QA.md	QA-agent rules for Ragger/Speculos testing and snapshot workflow.
.github/prompts/SKILL_LEDGER_DOC.md	Doc-agent rules for generating/maintaining APDU + design documentation.
.github/prompts/SKILL_LEDGER_REVIEW.md	Review-agent “gatekeeper” checklist and verdict format.
.github/prompts/README.md	Explains how to use the orchestrator + skills as a prompt workflow.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-02-17T17:24:50Z

+4.  **Provide Proof (MANDATORY):**
+    - After writing code, you **MUST execute** the Docker Build Command **yourself in the terminal**. Do NOT just display it to the user.
+    - **Start Clean:** Begin the command chain with `make clean`.
+    - **Filter Targets:** Read `[app].devices` from `ledger_app.toml` and generate build commands **ONLY** for the listed devices. Use the **Target → SDK Variable Mapping** table above to translate each device name to its `BOLOS_SDK` variable (e.g., `nanos+` → `$NANOSP_SDK`).
+    - **Check Compliance:** **MUST** end the command chain with the Rule Enforcer (`/opt/enforcer.sh`).
+    - **Enforcer Duration:** The enforcer runs `scan-build` for every target and can take **10-15 minutes**. Use a generous timeout (900s+). Consider running the compilation chain and enforcer as **two separate commands** to avoid timeout-induced SIGTERM.
+    - **Chain Commands:** Use `&&` to ensure the process stops on the first error.
+    - **Preserve Artifacts:** Stop after the last compilation step (Do not run a final clean).
+    - **Execution:** Use `run_in_terminal` (or equivalent tool) to launch the `docker run` command. Do NOT ask the user to run it manually.
+5.  **Iterate (The Fix Loop):**


This prompt requires using a run_in_terminal tool, but this repository doesn't define such a tool and common chat integrations (e.g., GitHub Copilot Chat) can't execute Docker commands. Consider rephrasing to be tool-agnostic (e.g., "run the following commands locally") or documenting the specific integration that provides terminal execution.

Copilot · 2026-02-17T17:24:51Z

+- **Description** — What the command does.
+- **Command table** — CLA, INS, P1, P2, Lc, CData with value + description for each field.
+- **Response table** — Field name, length, description. Always end with SW `0x9000`.
+- **Error Responses table** — Condition → SW code. Cover: wrong P1/P2 (`0x6A86`), wrong data length (`0x6D02`), invalid data (`0x6A80`), user rejection (`0x6985`).


The example status word for "wrong data length" is incorrect (0x6D02 is not a standard length error and doesn't match what this codebase uses via StatusWords/SWO_WRONG_DATA_LENGTH). Prefer referencing the codebase constant (e.g., Errors.SWO_WRONG_DATA_LENGTH) or the standard 0x6700 to avoid generating inaccurate APDU docs.

Suggested change

- **Error Responses table** — Condition → SW code. Cover: wrong P1/P2 (`0x6A86`), wrong data length (`0x6D02`), invalid data (`0x6A80`), user rejection (`0x6985`).

- **Error Responses table** — Condition → SW code. Cover: wrong P1/P2 (`0x6A86`), wrong data length (`0x6700`), invalid data (`0x6A80`), user rejection (`0x6985`).

Copilot · 2026-02-17T17:24:51Z

+1.  **Analyze the Artifacts:** Read the changes in `src/`, the test directory (from `[pytest.standalone].directory` in `ledger_app.toml`), and `doc/`.
+2.  **Evaluate Logic:**
+    - Ask: *"Does this C code solve the user request?"*
+    - Ask: *"Does the Python test prove the feature works?"*
+    - Ask: *"Is the Documentation accurate?"*
+3.  **Produce Verdict:**
+    - **FAIL:** Return a list of specific "Blockers" if ANY principle is violated. Explicitly state what must be fixed.
+    - **PASS:** Output `✅ READY TO MERGE` ONLY if the feature is production-ready.


markdownlint (MD030) requires exactly one space after an ordered-list marker. Update this routine list (and the ordered list inside the blockquote examples below) to use 1. instead of 1. to avoid lint failures.

Copilot · 2026-02-17T17:24:51Z

+    - **Code blocker** (security, logic, build) → Loop to **PHASE 1** (Dev).
+    - **Test blocker** (missing coverage, wrong assertion) → Loop to **PHASE 2** (QA).
+    - **Doc blocker** (wrong INS code, missing diagram) → Loop to **PHASE 3** (Doc).
+    You MUST pass the list of "Blockers" found by the Reviewer as high-priority constraints for the targeted phase.  


This line ends with trailing spaces (used for a Markdown hard line break). The trailing-whitespace pre-commit hook will strip these spaces, so the hard break won't be preserved. Prefer an explicit <br> (allowed by MD033) or reflow the paragraph to avoid relying on trailing spaces.

Suggested change

You MUST pass the list of "Blockers" found by the Reviewer as high-priority constraints for the targeted phase.

You MUST pass the list of "Blockers" found by the Reviewer as high-priority constraints for the targeted phase.<br>

Copilot · 2026-02-17T17:24:52Z

+1.  **Integration Tests (Ragger):** MANDATORY.
+2.  **Unit Tests (Host):** OPTIONAL (Recommended for complex logic).
+3.  **Fuzzing:** OPTIONAL (Recommended for security hardening).


markdownlint (MD030) requires exactly one space after an ordered-list marker. Update this ordered list (and the other ordered lists in this file) to use 1. instead of 1. to avoid lint failures.

Suggested change

1. **Integration Tests (Ragger):** MANDATORY.

2. **Unit Tests (Host):** OPTIONAL (Recommended for complex logic).

3. **Fuzzing:** OPTIONAL (Recommended for security hardening).

1. **Integration Tests (Ragger):** MANDATORY.

2. **Unit Tests (Host):** OPTIONAL (Recommended for complex logic).

3. **Fuzzing:** OPTIONAL (Recommended for security hardening).

Copilot · 2026-02-17T17:24:53Z

+5.  **Provide Proof (MANDATORY):**
+    - After writing tests, you **MUST execute** the Docker Test Command **yourself in the terminal**. Do NOT just display it to the user.
+    - **Step 1: Setup:** Activate venv and install requirements.
+    - **Step 2: Golden Run:** Run `pytest --golden_run` for **ALL** supported devices.
+    - **Step 3: Verification:** Run `pytest` (without golden_run flag) for **ALL** supported devices to confirm stability.
+    - **Chain Commands:** Use `&&` to ensure the process stops immediately if any step fails.
+    - **Execution:** Use `run_in_terminal` (or equivalent tool) to launch the `docker run` command. Do NOT ask the user to run it manually.
+6.  **Iterate (The Fix Loop):**


This prompt requires using a run_in_terminal tool to execute Docker commands, but the repo doesn't define such a tool and many target chat environments cannot run commands. Consider rewriting to be tool-agnostic (provide commands to run) or explicitly documenting the required tooling/integration.

Copilot · 2026-02-17T17:24:53Z

+1.  Open your AI Chat (or "Composer" / "Copilot Edits" mode).
+2.  **Reference the `@ORCHESTRATOR.md` file** in your context (e.g. `#file:ORCHESTRATOR.md`).
+3.  Type your high-level request.


markdownlint (MD030) requires exactly one space after an ordered-list marker. Update this ordered list (and other ordered lists in this file) to use 1. instead of 1. to avoid lint failures.

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

add orchestrated agent skills and workflow

f01ffb1

tdejoigny-ledger requested a review from Copilot February 17, 2026 17:18

Copilot started reviewing on behalf of tdejoigny-ledger February 17, 2026 17:18 View session

Copilot AI reviewed Feb 17, 2026

View reviewed changes

tdejoigny-ledger and others added 3 commits February 17, 2026 18:29

Update .github/prompts/SKILL_LEDGER_C.md

4b81d54

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Update .github/prompts/SKILL_LEDGER_DOC.md

1ddaa20

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Update .github/prompts/ORCHESTRATOR.md

4927426

Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

add orchestrated agent skills and workflow#195

add orchestrated agent skills and workflow#195
tdejoigny-ledger wants to merge 4 commits into
masterfrom
tdj/AI_agents

tdejoigny-ledger commented Feb 17, 2026

Uh oh!

codecov Bot commented Feb 17, 2026 •

edited

Loading

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Feb 17, 2026

Uh oh!

Copilot AI Feb 17, 2026

Uh oh!

Copilot AI Feb 17, 2026

Uh oh!

Copilot AI Feb 17, 2026

Uh oh!

Uh oh!

Copilot AI Feb 17, 2026

Uh oh!

Copilot AI Feb 17, 2026

Uh oh!

Uh oh!

Copilot AI Feb 17, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

	- Error Responses table — Condition → SW code. Cover: wrong P1/P2 (`0x6A86`), wrong data length (`0x6D02`), invalid data (`0x6A80`), user rejection (`0x6985`).
	- Error Responses table — Condition → SW code. Cover: wrong P1/P2 (`0x6A86`), wrong data length (`0x6700`), invalid data (`0x6A80`), user rejection (`0x6985`).

	You MUST pass the list of "Blockers" found by the Reviewer as high-priority constraints for the targeted phase.
	You MUST pass the list of "Blockers" found by the Reviewer as high-priority constraints for the targeted phase.<br>

Uh oh!

Conversation

tdejoigny-ledger commented Feb 17, 2026

Uh oh!

codecov Bot commented Feb 17, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Feb 17, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 17, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 17, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 17, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Copilot AI Feb 17, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 17, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Copilot AI Feb 17, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

codecov Bot commented Feb 17, 2026 •

edited

Loading