Approach
Scanned 8/31/2026
Install to Claude Code
npx -y skills add instrument-org/instrument --skill changelog --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Changelog?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/instrument-org-changelog)More formats (shields.io, HTML) on the badges page.
---
name: changelog
description: Approach
disable-model-invocation: true
---
## Approach
1. Use `git` to look for the most recent tagged non-beta version.
2. If the most recent commit doesn't have a tag, use it as the current version.
3. Find the previous version.
4. Inspect the content of any larger commits.
5. Generate a change log for that based on the style of a few of these example change logs down below.
6. Return the results in a markdown code block.
## Guidelines
- Don't include verbatim commit messages.
- Write copy for a general audience.
- Only include the "Features" and "Bug Fixes" sections.
## Example 1
```markdown
## Features
- **Chat mode** has landed! Switch between building apps and chatting with the agent.
- **Tasks page** with table view - delete, stop, and manage all your tasks in one place.
- Recent tasks displayed on new tab.
- Copy button for pre blocks in Markdown.
- Use new tasks page for eval viewing.
- Incorporate upstream error messages for multiple matches.
## Bug Fixes
- Full width assistant messages always, and copy button for single line code too.
- Support navigation API even on root pages like not found.
- Remove redundant app type in tab title.
- Remove more markdown in titles.
- Attempt to avoid race condition during task deletion due to session abort.
- Make command/ctrl + r work everywhere unless disabled.
- Disallow creation of session db during entire deletion process.
- Better handle ripgrep arguments.
- Animated stop icon in proper places.
## ⛵︎
_72 commits since v1.5.0_
More will follow.
```
## Example 2
```markdown
## 🧬 Features
- Evals have landed!
- A couple clicks and you can compare rich output from as many models and apps you desire.
- Built-in prompts for standard eval apps and some goofy ones too.
- Support for custom eval prompts.
- Dedicated runs page with bulk actions.
## 🩹 Bug Fixes
- Linux: Fixed DBus safe storage issues by setting up direct DBus communication.
- Workspace: Ensured runtime starts when agent session spawns to prevent missing dependencies.
- AI Gateway: Fixed handling of `:latest` style model names (e.g., `gpt-oss:latest`).
- Workspace: Blocked `pnpm dev`, `pnpm start`, and `pnpm run` variants since Quests already manages app lifecycle.
## Water Vehicle
<img width="512" height="512" alt="create_an_art_deco_abstract_painting_of_a_streamlined_boat_gliding_through_geometric_luxury_incorpo_2zoya69fyv6xzft55rlj_1" src="https://github.com/user-attachments/assets/51ee86dc-f5d8-4489-b9ba-fc308d932b90" />
_23 commits since v1.4.2_
```
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!