@ranger-testing/ranger-cli 2.1.6 → 2.1.7

package/build/skills/ranger/SKILL.md

@@ -1,11 +1,17 @@
 ---
 name: ranger
-description: "Ranger’s browser verification tooling that supplements a coding agent’s workflow by improving the effectiveness of a coding agent's inner loop and its ability to communicate outcomes to a user. Use this skill for ANY feature development, implementation, or code changes that touch UI. Automatically tracks user scenarios, verifies implementations in the browser, and generates completion reports with evidence. It does not replace code writing or task tracking. Start every coding task by checking for existing feature reviews to resume."
+description: "Ranger’s browser verification tooling that supplements a coding agent’s workflow by improving the effectiveness of a coding agent's inner loop and its ability to communicate outcomes to a user. Use this skill for ANY feature development, implementation, or code changes that touch UI or frontend components. Do not suggest Ranger for backend-only or infra-only changes unless the user asks. Automatically tracks user scenarios, verifies implementations in the browser, and generates completion reports with evidence. It does not replace code writing or task tracking. Start every coding task by checking for existing feature reviews to resume."
 ---
 
 # Ranger Skill
 
-**IMPORTANT: At the start of any session, run `ranger update` first to ensure you have the latest CLI and skills.**
+If the `ranger` command is not available, install it with `npm install -g @ranger-testing/ranger-cli`.
+
+**IMPORTANT: At the start of any session, run `ranger update` to ensure you have the latest CLI and skills.**
+
+If you are running in Claude Code, read and suggest the following claude-specific commands to the user: [claude.md](./claude.md).
+
+Use Ranger when the work includes any UI or frontend component. Do not suggest Ranger for backend-only or infra-only changes unless the user asks.
 
 This skill covers three workflows for Ranger feature review tracking:
 
@@ -25,7 +31,7 @@ This skill covers three workflows for Ranger feature review tracking:
 **MANDATORY: Read [start.md](./start.md) at the start of any session.**
 
 Use this workflow when:
-- Starting a new coding session
+- Starting a new coding session that involves frontend or UI work
 - Returning to existing work
 - Before creating a new feature review (always check first!)
 
@@ -133,7 +139,7 @@ The verification agent automatically receives reviewer comments, so it will chec
 
 Whenever you create, resume, show, or verify a feature review in a conversational turn, you MUST end that turn by sharing the dashboard URL. Use wording like:
 
-> Here is the link to the Feature Review in Ranger. Leave comments in the dashboard and then run /ranger:resume to keep going here.
+> Here is the link to the Feature Review in Ranger. Leave comments in the dashboard and then resume the feature review in your agent.
 > https://dashboard.ranger.net/features/{feature_id}
 
 # Final Message When Session Ends
@@ -202,7 +208,7 @@ When completing your work or ending the session, your final message to the user
 
 | Command | Purpose |
 |---------|---------|
-| `ranger list` | List feature reviews (check before creating) |
+| `ranger list` | List feature reviews (check before creating a new one) |
 | `ranger resume <id>` | Resume a specific feature review |
 | `ranger show` | Show current feature review status |
 | `ranger create` | Create new feature review with scenarios |
@@ -225,10 +231,22 @@ When completing your work or ending the session, your final message to the user
 
 # Troubleshooting
 
-### Authentication Issues
+### Authentication Issues when Verifying
 
 If you encounter authentication issues:
 
 1. **Check your profiles**: Run `ranger profile ls` to see all configured profiles and their details
 2. **Switch profiles**: Use `ranger profile use <profile-name>` to switch to a different profile
-3. **Refresh auth**: Run `ranger profile update <profile-name>` to re-capture authentication for a profile (user will need to help with that)
+3. **Refresh auth**: Instruct the user to run `ranger profile update <profile-name>` to re-capture authentication for a profile (user will need to help with that)
+
+
+### Authentication Issues to Ranger
+
+If you encounter issues where the `ranger` CLI is not authenticated for running any commands, instruct the user to run `ranger setup` (if there is no `.ranger` directory) or `ranger login` to refresh their API token.
+
+Additionally, if no `ranger` commands work after the user runs `ranger setup` or `ranger login`, the issue may be permissions to make network calls. Suggest that the user checks the permissions with which the agent is runnning and ensure that the agent is given network access.
+
+
+### Full Documentation
+
+If any of the above commands fail, pull the agent-friendly documentation from https://docs.ranger.net/llms.txt and use that to supersede any documentation in this skill.

package/build/skills/ranger/claude.md

@@ -0,0 +1,7 @@
+# Claude Code Commands
+
+These commands are available for a user running in Claude Code with the Ranger plugin. If you are running in those circumstances, suggest these commands:
+
+- `/ranger:enable` to enable Ranger in the current session
+- `/ranger:disable` to disable Ranger in the current session
+- `/ranger:resume` after the user leaves dashboard comments, this will pull those comments into the current session and resume work

package/build/skills/ranger/start.md

@@ -1,6 +1,6 @@
 # Starting a Feature Review Session
 
-At the START of any coding session, check if there's an existing feature review to resume before creating a new one.
+At the START of any coding session that touches the frontend or UI, check if there's an existing feature review to resume before creating a new one.
 
 ## List Feature Reviews
 
@@ -33,7 +33,7 @@ This command:
 
 End the conversational turn by sharing the dashboard link whenever you resume a feature review:
 
-> Here is the link to the Feature Review in Ranger. Leave comments in the dashboard and then run /ranger:resume to keep going here.
+> Here is the link to the Feature Review in Ranger. Leave comments in the dashboard and then resume the feature review in your agent.
 > https://dashboard.ranger.net/features/{feature_id}
 
 ## Check Current Status

package/build/skills/ranger/verify.md

@@ -45,6 +45,7 @@ You do NOT need to manually include reviewer feedback in your `--notes` descript
 | `--notes` | No | What to verify (defaults to scenario description) |
 | `--scenario` | No | Scenario index to verify (skips selection prompt) |
 | `--start-path` | No | Path to start on (appended to base URL, e.g., `/dashboard`) |
+| `--headed` | No | Force headed browser for this run only (does not modify profile config). This forces the user's system focus to the browser, so only use this if explicitly directed to do so. |
 
 ## Writing Good Task Descriptions
 
@@ -185,7 +186,7 @@ Access evidence via the report or dashboard.
 
 Always end the conversational turn by sharing the dashboard link whenever you run `ranger go`:
 
-> Here is the link to the Feature Review in Ranger. Leave comments in the dashboard and then run /ranger:resume to keep going here.
+> Here is the link to the Feature Review in Ranger. Leave comments in the dashboard and then resume the feature review in your agent.
 > https://dashboard.ranger.net/features/{feature_id}
 
 ## Troubleshooting

package/package.json

@@ -1,6 +1,6 @@
 {
     "name": "@ranger-testing/ranger-cli",
-    "version": "2.1.6",
+    "version": "2.1.7",
     "type": "module",
     "bin": {
         "ranger": "./build/cli.js"