[rameshvaranganti]

Pull Request Checklist

• I have read and followed the CONTRIBUTING.md guidelines.
• I have read and followed the Guidance for submissions involving paid services.
• My contribution adds a new instruction, prompt, agent, skill, workflow, or canvas extension file in the correct directory.
• The file follows the required naming convention.
• The content is clearly structured and follows the example format.
• I have tested my instructions, prompt, agent, skill, workflow, or canvas extension with GitHub Copilot.
• I have run npm start and verified that README.md is up to date.
• I am targeting the main branch for this pull request.

Description

Summary

Adds the sap-cloud-integration skill (SAP Cloud Integration / CPI runtime knowledge) and the SAP Integration Suite Copilot agent, which delegates CPI questions to it.

This reworks the submission after the feedback on #6364, which said the earlier version was "a short persona prompt and broad checklist" with no example of what it gets right that a capable model would miss. The new version is a skill with four reference files of CPI runtime behavior:

• Groovy script API: streaming body access, MPL properties vs. custom header properties vs. gated attachments, SecureStoreService, ValueMappingApi
• Header vs. exchange property scope: ProcessDirect, JMS (Transfer Exchange Properties, 4 MB limit), HTTP receiver Request Headers (none by default), Allowed Header(s), SAP_* / Camel* names the runtime reads
• Exception Subprocess outcomes: End → Completed, Error End → Failed, Escalation End → Escalated; SAP_MessageProcessingLogCustomStatus; CamelExceptionCaught with AhcOperationFailedException / OsciException; JMS retry with SAPJMSRetries (Non-Exclusive queues)
• MPL error → cause → check: SFTP, HTTP/OAuth/certificates (including SAP_DisableMonitoringHeaders for unexplained HTTP 400), Cloud Connector, IDoc, OData, SOAP, AS2

Every behavior claim is cross-checked against the SAP Help Portal source in SAP-docs/btp-integration-suite. references/sap-documentation-index.md links each topic to the SAP page it comes from. Author: SAP integration architect with 20+ years on SAP PI/PO and Integration Suite.

Evaluation: same model, with and without the skill

Environment: VS Code 1.145.3, GitHub Copilot Chat 0.33.1, Agent mode, model Copilot Fusion (auto-routed to gpt-6-luna). Each prompt ran in a new chat, first in an empty workspace (plain), then in a workspace containing only .github/skills/sap-cloud-integration (skill). Screenshots below.

#	Prompt	Without skill	With skill
1	"In the Exception Subprocess I need the receiver's HTTP status and error body."	❌ Checks for org.apache.camel.component.http.HttpOperationFailedException with instanceof. CPI's HTTP receiver throws org.apache.camel.component.ahc.AhcOperationFailedException (SAP script example), so the script never matches and always throws.	✅ Reads CamelExceptionCaught, compares by class name to AhcOperationFailedException as SAP's example does, extracts getStatusCode(), getStatusText(), getResponseBody(), and notes the OData V2 difference. Copilot shows it read error-handling-reliability.md.
2	"Write a Groovy script that logs the full payload for every message to the MPL."	⚠️ Attaches the payload unconditionally to every message's MPL, with a general caution only.	✅ Same script, but the attachment is gated by a property fed from an externalized LOG_PAYLOAD flag, with a tenant test (flag off → no attachment) and storage/PII guidance.
3	"My Exception Subprocess … ends with an End event. Support says failed messages don't show up as failed. Why?"	⚠️ Correct main fix (Error End).	✅ Also: Escalation End for status Escalated, CamelHttpResponseCode so synchronous callers don't get HTTP 200, and SAP_MessageProcessingLogCustomStatus when the error is handled on purpose, so support can filter by Custom Status.

Where the skill did not add much: for "property lost over JMS", both answers proposed copying it into a header (the skill only added the Transfer Exchange Properties caution). For "IDoc status 56", the plain answer was already good. That test exposed an overly narrow line in the skill, which I corrected before this submission.

Screenshots: with and without the skill (click to expand)

Without skill — #1 HTTP status in Exception Subprocess

With skill — #1 HTTP status in Exception Subprocess

Without skill — #2 Payload logging

With skill — #2 Payload logging

Without skill — #3 End event in Exception Subprocess

With skill — #3 End event in Exception Subprocess

Without skill — Property lost over JMS

With skill — Property lost over JMS

Without skill — IDoc status 56

With skill — IDoc status 56 (before the IDoc correction)

Changes since the first review

• Agent prompt no longer contains an author line; generated code carries no attribution.
• Status note requires a confirmed successful tenant test; deployment alone is not enough.
• Agent reduced to a thin entry point; CPI knowledge moved into the skill.

[github-actions]

🔍 Vally Lint Results

⚠️ Warnings or advisories found

Scope	Checked
Skills	1
Agents	1
Total	2

Severity	Count
❌ Errors	0
⚠️ Warnings	0
ℹ️ Advisories	1

Summary

Level	Finding
ℹ️	Vally currently lints SKILL.md content. Agent files were detected but skipped:

Full linter output

### Linting skills/sap-cloud-integration
✅ sap-cloud-integration (2/2 checks passed)
    ✓ [spec-compliance] All 1 skill(s) are spec-compliant.
        ✓ spec-compliance: All spec checks passed.
    ✓ [valid-refs] All file references across 1 skill(s) are valid.
        ✓ valid-refs: All file references resolve to existing files within the skill directory.

1 skill(s) linted, 1 passed

### Agent files detected (not linted by vally)
ℹ️ Vally currently lints SKILL.md content. Agent files were detected but skipped:
agents/sap-integration-suite-copilot.agent.md

[github-actions]

🔒 PR Risk Scan Results

Scanned 8 changed file(s).

Severity	Count
🔴 High	0
🟠 Medium	0
ℹ️ Info	0

✅ No matching risk patterns were detected in changed files.

This is an automated soft-gate report. Findings indicate review targets and do not block merge by themselves.

[github-actions]

🚦 Submission status: 👀 Ready for review

Risk tier: merge-risk:medium — New or substantially changed resource without privileged execution
Required to merge: passing submission-gate checks plus 1 approval from reviewers with write access.

Why this tier

• Adds 8 file(s), e.g. agents/sap-integration-suite-copilot.agent.md
• Changes 499 lines (low-risk limit is 40)

Automated checks

Check	Status	Details
Line endings	✅ Passed	Passed · logs
Spelling	✅ Passed	Passed · logs
Generated README consistency	✅ Passed	Passed · logs
Skill validation	✅ Passed	Passed · logs
Skill lint (vally)	✅ Passed	Passed · logs
Risk scan	✅ Passed	Passed · logs
Contributor reputation	✅ Passed	Passed · logs
Duplicate resource scan	✅ Passed	Passed · logs
PR quality signal	⏭️ Skipped	Skipped by its workflow · logs

Review

• Approvals: 0/1
• Assigned reviewer: not assigned yet — comment /request-review to ask for one
• Review target date: not set
• Still needed: 1 more approval(s)
• No staffed reviewer pool owns these files yet; any reviewer with write access counts as the domain reviewer.

Commands

Command	Who	What it does
/rerun-checks	PR author, maintainers	Re-runs failed or incomplete checks and re-evaluates this gate
/request-review	PR author, maintainers	Asks the review rotation to assign a reviewer (adds needs-reviewer)

_{Updated for e83c4b3 · gate run · This comment is maintained automatically — see submission gate docs.}

[Copilot]

🟡 Changes recommended

The agent requires generated scripts to credit its author and promote their website, creating misleading provenance in unrelated user code.

2 open issues

• Remove mandatory attribution from generated scripts · New
• Add a focused CPI workflow with runtime constraints and examples · New

What changed in this PR

Adds an SAP Integration Suite agent to the repository’s Copilot agent collection, with guidance for integration development, review, and troubleshooting.

Changes:

• Defines SAP-focused scope, working practices, accuracy rules, and handoffs for reviews and test generation.
• Adds the agent to the catalog with VS Code and VS Code Insiders installation links.

File	Description
docs/​README.agents.md	Lists the new agent and installation links.
agents/​sap-integration-suite-copilot.agent.md	Defines the SAP Integration Suite agent and its workflows.

_{Review effort: Balanced (auto)}

[Copilot]

🔵 Needs a closer look

The agent’s SAP-specific guidance remains too limited, with generic capability lists and checklists needing replacement by a concrete CPI workflow, runtime constraints, and expected results.

No open issues 🎉

1 previously missed issue

In code that hasn't changed since last review

Replace the PR template with Copilot evaluation evidence

agents/​sap-integration-suite-copilot.agent.md:17

The PR description is still an unfilled template and provides no realistic usage scenario or evaluation results, so maintainers cannot assess whether this agent works in Copilot as intended. The repository's review criteria ask for validation evidence; please document a representative CPI request, the Copilot environment, and expected versus observed results, including the review and test-case handoffs. A manual evaluation is sufficient; tenant deployment is not required.

2 resolved since last review

• Remove mandatory attribution from generated scripts
• Add a focused CPI workflow with runtime constraints and examples

_{Review effort: Balanced (auto)}

[rameshvaranganti]

please review

[rameshvaranganti]

Certificate-trust guidance tightened: the TLS test’s presented chain is now treated as unauthenticated input. The skill requires obtaining CA certificates from an independent source and matching SHA-256 fingerprints before upload (including before using the TLS test’s Add to keystore), trusting CAs rather than leaf certificates, and never disabling certificate validation.

[Copilot]

🔵 Needs a closer look

Human assessment is needed because line-based processing and unbounded payload logging can allocate entire large payloads and exhaust worker memory.

No open issues 🎉

_{Review effort: Balanced (auto)}

[rameshvaranganti]

Memory guidance tightened per SAP’s Script API docs: scripts now check message.getBodySize() before any getBody(String) (SAP warns String/byte[] conversion can cause OOM), payload attachments are both flag-gated and size-capped, the 1 GB/24 h MPL attachment limit is documented, and line-by-line reading is limited to line-delimited formats. Sources: [Read and Modify Message Header, Message Body, and Exchange Properties]

[Copilot]

🟡 Changes recommended

The payload-size guard accepts getBodySize() values of -1, allowing unknown-size bodies to bypass the logging cap and potentially attach arbitrarily large payloads.

1 open issue

• Skip payload attachment when body size is unknown · New

_{Review effort: Balanced (auto)}

[rameshvaranganti]

resolved

[Copilot]

🔵 Needs a closer look

Human assessment is needed to evaluate the sensitive-data exposure risk from persisting raw payloads in log attachments without enforced redaction.

No open issues 🎉

1 resolved since last review

• Skip payload attachment when body size is unknown

_{Review effort: Balanced (auto)}

[rameshvaranganti]

please review

[aaronpowell]

Thanks for revising the SAP Integration Suite contribution. The updated status rule now distinguishes deployment from a confirmed tenant test, and the PR description documents three manual GitHub Copilot comparisons with observed results. The runtime guidance on Exception Subprocess status, AhcOperationFailedException, and JMS retry behavior is backed by linked SAP sources. Please continue to keep the tenant-testing status explicit when this agent generates answers.