Julien Ramboz avatar

Julien Ramboz

Community

@ramboz · Bay Area, USA

9Followers
|
66Public Repos
|
34Published Skills

Tech Lead at Adobe. Building AI-native engineering workflows, deterministic validation systems, and developer tools. Creator of jig and servo.

Agent Skills by Julien Ramboz

Showing 34 vetted skills indexed across 2 GitHub repositories.

rambozramboz

servo:oracle-hook

Automate installation, uninstallation, and status checks of the meta-judge Stop hook in Claude Code.

Community
Advanced
rambozramboz

eval-authoring

Automates creation of rubrics, reference sets, and evaluation components for text-based acceptance criteria.

Community
Advanced
rambozramboz

spec-oracle

Compile project specifications into executable evaluation overlays for automated code checks.

Community
Advanced
rambozramboz

servo:agent-loop

Automate iterative testing and optimization over codebases with Claude Code agents.

Community
Advanced
rambozramboz

servo:scaffold-init

Automates Servo project setup by detecting signals and installing an oracle.sh evaluation script.

Community
Advanced
rambozramboz

servo:quality-gate

Execute predefined scripts to evaluate software project quality with exit codes.

Community
Advanced
rambozramboz

servo:heartbeat

Monitors CI failures and open issues while scheduling evaluations with cost ceiling controls.

Community
Intermediate
rambozramboz

design-eval

Capture screenshots, compare UI against design mockups, and generate rubric-based fidelity scores.

Community
Advanced
rambozramboz

servo:edd-suitability

Analyze engineering specs for Evaluation-Driven Development suitability.

Community
Advanced
rambozramboz

servo:execution-planner

Generate the ADR-0016 execution plan for Servo Compile→Run handoffs.

Community
Intermediate
rambozramboz

oracle-hook

Manage installation, uninstallation, and status of Claude Code's 'Stop' hook.

Community
Intermediate
rambozramboz

agent-loop

Run iterative agent loops with guardrails and gate.py scoring.

Community
Advanced
rambozramboz

content-fidelity

Evaluate generated text against predefined rubrics using the Anthropic Messages API.

Community
Intermediate
rambozramboz

quality-gate

Run automated scripts to evaluate project quality against defined criteria.

Community
Advanced
rambozramboz

heartbeat

Collect project signals and dispatch findings through the Servo oracle-gated loop.

Community
Advanced
rambozramboz

edd-suitability

Evaluate engineering specifications for Evaluation-Driven Development suitability and missing evidence.

Community
Advanced
rambozramboz

execution-planner

Assemble the Servo Compile execution plan from compiled inputs into a durable handoff artifact.

Community
Advanced
rambozramboz
5

memory-sync

Persist session-derived context to the memory layer in CLAUDE.md.

Community
Advanced
rambozramboz
5

explain

Translate jig vocabulary and dense artifacts into plain language.

Community
Intermediate
rambozramboz
5

independent-review

Generate standardized reviewer prompts for multiple review passes against specs.

Community
Advanced
rambozramboz
5

clarify

Audits a spec.md or slice file for ambiguities across six defined categories and generates prioritized questions.

Community
Advanced
rambozramboz
5

contracts

Recommend canonical contract schemas and validate them with ecosystem tools.

Community
Intermediate
rambozramboz
5

scaffold-init

Initialize an AI-native workspace scaffold by copying templates and injecting runtime paths.

Community
Advanced
rambozramboz
5

tdd-loop

Automate the red-green-refactor loop with pytest, vitest, or jest runners.

Community
Intermediate