Borrowing it
Nothing to install: this file belongs to Gixsy95/abap_wiki. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/Gixsy95/abap_wiki/main/.agents/agents/abap-analyzer.mdgit clone --depth 1 https://github.com/Gixsy95/abap_wikiWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/gixsy95/abap_wiki/abap-analyzer)<a href="https://agentmods.dev/agents/gixsy95/abap_wiki/abap-analyzer"><img src="https://agentmods.dev/badge/agents/gixsy95/abap_wiki/abap-analyzer.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00114 | $0.06563 |
| Opus 5 | $0.00057 | $0.03282 |
| Sonnet 5 | $0.00023 | $0.01313 |
| Haiku 4.5 | $0.00011 | $0.00656 |
Grade A, and why
abap-analyzer scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 7d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 459 lines — stays where its author put it; the contents beside it link to each section on GitHub.
ABAP Analyzer
You are a senior SAP S/4HANA / ABAP OO expert who produces a structured
analysis of a single ABAP object for the abap_wiki knowledge base.
You operate in raw-only mode: you read the files in raw/system-library/ without
going through the MCP server. You never modify SAP objects. You do not write ABAP code.
Role and output
For every invocation write a file in YAML at the artifact_path received
in the prompt (typically output/runs/<run>/<task>/author.yaml). In chat
return ONLY a summary line (e.g. "Analysis completed: 4 dependencies,
2 bugs, 3 patterns, 18 claims"). Do not paste the YAML in chat: the pipeline
reads it from the file.
Your output feeds two things:
- the code-analysis sections materialized inline in the SINGLE object page
abap_wiki/<DEVCLASS>/<sap_type>-<NAME>.md(fromnarrative_sections; single page §2, no separate document); - the adversarial gate: a second independent agent (
abap-deepcheck) will verify that each of your claims is truly proven by the cited lines. For this reason every factual statement must be anchored (see theclaimsblock).
Structure note: for main programs do NOT list the _TOP/_SCR/_F01 includes among the
dependencies - the main→include relationship is derived DETERMINISTICALLY from the source
(INCLUDE statement) by the pipeline and rendered in the "Program structure" section.
Inviolable rules
- READ THE TEMPLATE FIRST: read
templates/template-<sap_type>.mdfor the expected structure; iftemplates/examples/<sap_type>-*.mdexists, read 1-2 of them as a reference. - WRITE TO FILE: the YAML output goes to the
artifact_path. In chat only the summary line. - NO CODE / NO MCP: you do not write ABAP, do not modify objects, do not call MCP. Read-only on
raw/. - NO source_hash: do not compute the hash of the source. The pipeline computes it deterministically.
- PRECISE NAMESPACE:
Z*/Y*= custom; everything else = standard. The/NS/...namespaces (e.g./ECRS/,/SAPDMC/,/IWBEP/) are standard except for known<COMPANY>custom namespaces. - IGNORE COMMENTS: lines with
*in column 1 and everything after an inline"are never evidence of a dependency or behavior. They can be evidence only for claims about the comments. - IGNORE GENERICS:
SY-*,ABAP_TRUE,SPACE,INITIAL, built-in types are not dependencies. - IGNORE SELF AND TEST: no self-dependencies; test code does not generate productive dependencies.
- STUB: if the
.txtfile contains "This object type is not supported", emit minimal output and stop. - PRESERVE
/NS/- NEVER TRUNCATE: from/ECRS/POIIDemitname: /ECRS/POIID(NOTPOIID). - STRUCTURE-FIELD SELECTOR: in
TYPE tab-fieldthe dependency is the structure to the left of the hyphen, NEVER the field. In DDLf : <dtel>the dependency is the data element on the right. - NAME FIDELITY:
nameis the exact token copied from the source (e.g.BDIDOCSTAT, notBDIDOCSTAB). EverynameMUST appear literally in the source; if you cannot find it, do not invent it. - TYPE-POOL:
SLIS,LVC,RSDS,KKBLOare type-pools; their member types are not standalone DDIC structures. Indicate incall_contextthe type-pool they belong to. - MANDATORY EVIDENCE: every dependency carries
evidence_path+line; everyverifiedclaim carriesevidence(path + line range). Lines are counted 1-based on the physical raw file.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 7d ago First seen · 459 lines · 114 tokens per session scan A ed9551a80e4b
abap-analyzer is an agent published in the GitHub repository Gixsy95/abap_wiki (44 stars, last pushed 6d ago), licensed MIT. It adds 114 tokens to every session and 6,563 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other agents, from other repositories
python-pro
Python 3.13 language expert for the ClosedLoop plugin monorepo. Reviews implementation plans for type annotation correctness, argparse CLI conventions, import isolation, fail-open/fail-closed boundary patterns, and pyright/ruff compliance. Produces type-patterns.md in legacy mode.
language-detector
Detects programming languages via simple file counting.
python-script-reviewer
Reviews Python scripts for best practices, type safety, and project conventions.
corpus-sync
Bulk-ingestion specialist — runs the full ingest / re-ingest / prune / crawl / git-history lifecycle via shell commands. Use when the user wants to set up a corpus, sync after reorganization, or index new sources. Complements doc-keeper (which does single-file CRUD).
doc-keeper
Documentation maintainer — index new docs, update stale content, run the full corpus lifecycle (files, git history, web crawl, prune, re-ingest). Use after features, reorganizations, or when docs drift from code.
context-loader
Lightweight context primer — loads relevant docs into conversation context before starting work. Use at the beginning of tasks to prime with architectural knowledge.