How do I install the Sap Code Agents Analysis skill?

Run npx skillmds@latest add williamcorrea23/sap-code-agents-analysis in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Sap Code Agents Analysis safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Sap Code Agents Analysis?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Sap Code Agents Analysis free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Sap Code Agents Analysis?

williamcorrea23 (@williamcorrea23) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Sap Code Agents Analysis by williamcorrea23

Name: Sap Code Agents Analysis
Author: williamcorrea23

Sap Code Agents Analysis

Evidence-backed analysis of custom ABAP codebases, including object search, dependency and where-used analysis, usage evidence, migration findings, retirement candidates, and technical documentation. Adapted from jonathanbernsteiner/sap-code-agents.

File contents

SAP Code Agents Analysis

Use the evidence-first method adapted from jonathanbernsteiner/sap-code-agents.

Routing

Discover objects with the registered ADT/code-search capability.
Read the complete source of every object used as evidence.
Resolve callers and dependencies with where-used and repository search.
Use runtime usage data only when an authorized source is available; otherwise label it unavailable.
Produce findings that cite object names and source line evidence. Never convert an inference into a fact.

For a local abapGit checkout, search and read the files directly. For a live system, prefer the router's ADT providers. Do not ask for credentials in chat and do not bypass the configured MCP approval policy.

Deliverables

Functional purpose per object, supported by retrieved source.
Dependency and data-access map.
Migration risks with explicit evidence tiers.
Retirement candidates separated into supported, inferred, and unresolved.
Open questions for missing source or usage telemetry.

Evidence rules

Attach at least one verifiable object/source reference to each material claim.
Report parser or retrieval failures instead of silently omitting objects.
Keep system/workspace scope on every query; never mix evidence across systems.
Do not claim that unused code is safe to retire from static analysis alone.

Upstream reference

The pinned snapshot is cataloged as jonathanbernsteiner-sap-code-agents under bundled/skills/. Its architecture and evidence model are described in its README.md; its application is not an MCP server.

Verification

Confirm that object counts reconcile, all cited objects were retrieved, unresolved items are listed, and conclusions can be traced to source or authorized usage evidence.

---name: sap-code-agents-analysisdescription: Evidence-backed analysis of custom ABAP codebases, including object search, dependency and where-used analysis, usage evidence, migration findings, retirement candidates, and technical documentation. Adapted from jonathanbernsteiner/sap-code-agents.---# SAP Code Agents AnalysisUse the evidence-first method adapted from `jonathanbernsteiner/sap-code-agents`.## Routing1. Discover objects with the registered ADT/code-search capability.2. Read the complete source of every object used as evidence.3. Resolve callers and dependencies with where-used and repository search.4. Use runtime usage data only when an authorized source is available; otherwise label it unavailable.5. Produce findings that cite object names and source line evidence. Never convert an inference into a fact.For a local abapGit checkout, search and read the files directly. For a live system, prefer the router's ADT providers. Do not ask for credentials in chat and do not bypass the configured MCP approval policy.## Deliverables- Functional purpose per object, supported by retrieved source.- Dependency and data-access map.- Migration risks with explicit evidence tiers.- Retirement candidates separated into `supported`, `inferred`, and `unresolved`.- Open questions for missing source or usage telemetry.## Evidence rules- Attach at least one verifiable object/source reference to each material claim.- Report parser or retrieval failures instead of silently omitting objects.- Keep system/workspace scope on every query; never mix evidence across systems.- Do not claim that unused code is safe to retire from static analysis alone.## Upstream referenceThe pinned snapshot is cataloged as `jonathanbernsteiner-sap-code-agents` under `bundled/skills/`. Its architecture and evidence model are described in its `README.md`; its application is not an MCP server.## VerificationConfirm that object counts reconcile, all cited objects were retrieved, unresolved items are listed, and conclusions can be traced to source or authorized usage evidence.

Frequently asked questions