fix(harness): clarify skill shell path handling

[ZHCHJ888]

AgentScope-Java Version

2.0.3-SNAPSHOT

Description

Fixes #2788.

Skill execution guidance previously referred to a non-existent execute_shell_command tool and could lead agents to pass an absolute <files-root> path as working_directory. However, working_directory intentionally accepts only workspace-relative paths, so this caused skill script execution to be rejected.

This PR:

• Uses the actual execute tool name in the generated skill prompt.
• Clarifies that absolute skill paths belong in command, while working_directory must remain workspace-relative or be omitted.
• Improves the tool schema and validation error with recovery guidance without relaxing the workspace security boundary.
• Adds regression coverage for the generated prompt, rejected absolute working directories, and the exposed tool schema.
• Updates the English and Chinese skill documentation and examples.

How to test:

mvn -pl agentscope-harness -am -Dtest=SkillRuntimeTest,ShellExecuteToolTest -Dsurefire.failIfNoSpecifiedTests=false test
mvn -pl agentscope-harness -am -Dtest='io.agentscope.harness.**.*Test' -Dsurefire.failIfNoSpecifiedTests=false clean verify

Local results: 34 targeted tests passed. The clean Harness verification passed 989 tests with 6 environment-dependent tests skipped on Windows (5 Windows-disabled shell integration tests and 1 POSIX-permissions assumption).

Checklist

Please check the following items before code is ready to be reviewed.

• Code has been formatted with mvn spotless:apply
• All tests are passing (mvn test)
• Javadoc comments are complete and follow project conventions
• Related documentation has been updated (e.g. links, examples, etc.)
• Code is ready for review

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!