Add GitAgent Protocol support (agent.yaml + SOUL.md)#758
Conversation
|
Welcome to career-ops, @computer-agent! Thanks for your first PR. A few things to know:
We'll review your PR soon. Join our Discord if you have questions. |
📝 WalkthroughWalkthroughThis PR introduces the career-ops agent through two complementary files: SOUL.md documents the agent's behavioral identity, principles, capabilities, and operational constraints; agent.yaml configures the agent's metadata, model, skills, runtime parameters, and compliance controls. ChangesCareer-Ops Agent Definition
🎯 1 (Trivial) | ⏱️ ~3 minutes 🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 2
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Inline comments:
In `@agent.yaml`:
- Around line 10-14: The descriptive skill list text is inconsistent with the
actual skills declared in the skills block: the description claims "14 skill
modes" and lists names like evaluate/compare/contact while the skills block
contains 15 keys (including oferta/ofertas/contacto and interview-prep). Update
the description paragraph to exactly match the skills block: correct the count
to 15 (or remove the extra skill from the skills block if that was unintended),
and replace the enumerated names with the canonical skill keys used in the
manifest (e.g., ensure evaluate, compare, pipeline, PDF, scan, batch, deep,
training, project, tracker, apply, patterns, followup and interview-prep or the
oferta/ofertas/contacto variants match the skills block) so the names and count
are consistent with the skills list.
- Around line 24-39: The skills list in agent.yaml is missing the documented
"latex" mode referenced in SOUL.md; update the skills sequence under the
"skills:" key to include "latex" (lowercase) so the LaTeX export mode is
registered and discoverable by GAP runners and the registry, or alternatively
remove the "latex" entry from SOUL.md if the mode was intentionally deprecated;
locate the "skills" array in agent.yaml and add the "latex" token consistent
with other skill strings.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: Path: .coderabbit.yaml
Review profile: ASSERTIVE
Plan: Pro
Run ID: e2ff9912-ceeb-4bb1-b889-2f1a465b2136
📒 Files selected for processing (2)
SOUL.mdagent.yaml
| preparation. Supports 14 skill modes (evaluate, compare, pipeline, PDF, scan, | ||
| batch, contact, deep, training, project, tracker, apply, patterns, followup) | ||
| and multi-language operation (EN/DE/FR/JA). Designed for quality over | ||
| quantity: the agent will not submit applications without explicit user | ||
| approval. |
There was a problem hiding this comment.
Description mode list is inconsistent with the skills block.
The description states "Supports 14 skill modes" and enumerates 14 entries, but the skills list below declares 15. The names also diverge (evaluate/compare/contact here vs. oferta/ofertas/contacto in skills), and interview-prep is missing from this enumeration. Since this manifest is intended for the open registry, align the count and naming with the actual skills keys.
🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@agent.yaml` around lines 10 - 14, The descriptive skill list text is
inconsistent with the actual skills declared in the skills block: the
description claims "14 skill modes" and lists names like
evaluate/compare/contact while the skills block contains 15 keys (including
oferta/ofertas/contacto and interview-prep). Update the description paragraph to
exactly match the skills block: correct the count to 15 (or remove the extra
skill from the skills block if that was unintended), and replace the enumerated
names with the canonical skill keys used in the manifest (e.g., ensure evaluate,
compare, pipeline, PDF, scan, batch, deep, training, project, tracker, apply,
patterns, followup and interview-prep or the oferta/ofertas/contacto variants
match the skills block) so the names and count are consistent with the skills
list.
| skills: | ||
| - oferta | ||
| - ofertas | ||
| - pipeline | ||
| - scan | ||
| - batch | ||
| - contacto | ||
| - deep | ||
| - interview-prep | ||
| - training | ||
| - project | ||
| - tracker | ||
| - apply | ||
| - patterns | ||
| - followup |
There was a problem hiding this comment.
Missing latex skill — documented capability not registered.
SOUL.md (capabilities table, Line 38) documents a latex mode ("Export your CV as a polished LaTeX / Overleaf .tex file"), but it is not present in this skills list. As a result, the LaTeX export mode would not be advertised/invokable on GAP runners or in the registry, diverging from the documented behavior. Add it (or remove it from SOUL.md if intentionally dropped).
🔧 Proposed fix to register the documented mode
skills:
- oferta
- ofertas
- pipeline
- pdf
+ - latex
- scan
- batch
- contacto🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
In `@agent.yaml` around lines 24 - 39, The skills list in agent.yaml is missing
the documented "latex" mode referenced in SOUL.md; update the skills sequence
under the "skills:" key to include "latex" (lowercase) so the LaTeX export mode
is registered and discoverable by GAP runners and the registry, or alternatively
remove the "latex" entry from SOUL.md if the mode was intentionally deprecated;
locate the "skills" array in agent.yaml and add the "latex" token consistent
with other skill strings.
1 participant
Hi! This PR proposes GitAgent Protocol (GAP) support for career-ops — a small, open standard for portable AI agents (https://gitagent.sh).
Career-ops is already an excellent Claude Code agent. These two files are a lightweight, non-invasive addition that lets it also run on any GAP-compatible runtime and be discovered in the open registry.
What this adds (nothing else changes):
agent.yaml— standard manifest capturing the name, version, model (anthropic:claude-sonnet-4-6), 15 named skill modes, runtime limits, and thehuman_in_the_loop: alwayscompliance flag (faithfully reflecting career-ops' own rule that it never submits without user approval)SOUL.md— career-ops' persona and instructions distilled into the standard soul file, adapted fromCLAUDE.mdWith these two files, career-ops can run on any GAP-compatible runner and be listed in the open registry at https://registry.gitagent.sh. Totally optional to accept — feel free to tweak or close. Thanks for building career-ops in the open! 🦀
⭐ If GAP looks useful to you, the project lives at https://github.com/open-gitagent/opengap — a star helps more maintainers find the standard.
Summary by CodeRabbit
Release Notes
Documentation
Chores