Borrowing it
Nothing to install: this file belongs to palimkarakshay/abap-mcp. 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/palimkarakshay/abap-mcp/main/AGENTS.mdgit clone --depth 1 https://github.com/palimkarakshay/abap-mcpWrote 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/instructions/palimkarakshay/abap-mcp/agents-md)<a href="https://agentmods.dev/instructions/palimkarakshay/abap-mcp/agents-md"><img src="https://agentmods.dev/badge/instructions/palimkarakshay/abap-mcp/agents-md/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/instructions/palimkarakshay/abap-mcp/agents-md"><img src="https://agentmods.dev/badge/instructions/palimkarakshay/abap-mcp/agents-md.svg" alt="Reviewed on agentmods" width="80" 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 | $0.01934 | $0.01934 |
| Opus 5 | $0.00967 | $0.00967 |
| Sonnet 5 | $0.00387 | $0.00387 |
| Haiku 4.5 | $0.00193 | $0.00193 |
Grade A, and why
abap-mcp AGENTS.md 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 8d 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 — 92 lines — stays where its author put it; the contents beside it link to each section on GitHub.
abap-mcp
MCP server for SAP ABAP: offline static analysis (abaplint), ABAP Cloud / Clean Core
readiness checks, phased migration planning, released-API lookup against SAP's bundled
Cloudification snapshot, RAP managed-BO scaffolding, plus 4 guided-workflow MCP prompts
(abap-from-spec / abap-review / abap-mentor / abap-migration-plan). No SAP system, no credentials, and no
user-filesystem — text in, structured JSON out (package-bundled data assets excepted; see
invariants). The analysis engine makes no outbound network calls; stdio is local, while the
optional HTTP transport accepts inbound MCP requests. Lumivara product line: SAP.
Public MIT repo (github.com/palimkarakshay/abap-mcp); npm package abap-mcp (bin + library).
Package manager: npm — Node >= 20
Commands (authoritative)
npm installnpm run check— typecheck + vitest (218 tests) + build + routing eval, in order = the CI gatenpm run typecheck/npm test/npm run build— the individual stepsnpm run build—tsc && node scripts/copy-data.mjs(tsc does NOT copy the bundled.json)node scripts/build-released-api-index.mjs— dev-only, refreshes the SAP released-API snapshot (the only script that hits the network; never runs at serve time)node dist/cli.js— run the stdio MCP servernode dist/http.js— guarded Streamable HTTP (127.0.0.1:3000/mcpby default)npm run eval:routing— deterministic host-routing gate over exact MCP metadata- Inspect:
npx @modelcontextprotocol/inspector --cli node dist/cli.js --method tools/list - Description lint (optional, cross-repo; mcp-kit is NOT currently checked out locally — clone
palimkarakshay/mcp-kitfirst):pnpm --filter @mcp-kit/lint run lint -- --root ~/projects/abap-mcp --threshold 90. The in-repo rubric tests inserver.test.tsenforce the same discipline on every tool and prompt.
Layout
src/abap.tools.ts— the 13 ToolSpecs (lint_abap [+focus rule packs], fix_abap [deterministic abaplint auto-fixes, batch-verified], check_cloud_readiness [+A–D grade +rewrite recipes], plan_cloud_migration [phased backlog +examples], compare_abap, scaffold_rap_bo, scaffold_abap_unit [failing-by-default test harness], get_object_dependencies [tiered graph +released flags +mermaid], check_released_api, list_abap_rules, explain_abap_rule, format_abap, get_abap_outline [+mermaid]). The registry exporttoolsis what @mcp-kit/lint discovers.src/prompts.ts— the 4 PromptSpecs (abap-from-spec [spec→validated code, the blank-page killer], abap-review, abap-mentor, abap-migration-plan): the packaged consultant workflows, rubric-tested like tools; they steer agents to real tools and repeat the honesty boundaries. Mirrored as Codex plugin skills inplugins/abap-mcp/skills/— keep the two in lockstep.src/abap/— engine.ts (Registry wrapper, filename inference, caps, focus tag filter, AST object-reference extraction) · readiness.ts (dual-parse diff + released-API cross-check + density-banded A–D grade) · plan.ts (deterministic phased-backlog arrangement of a ReadinessReport: stage playbook per category, S/M/L bands, exit criteria — no new judgments) · compare.ts (before/after: content-matched finding diff + grade movement + structure changes) · fix.ts (deterministic auto-fix: Issue.getDefaultFix edits via abaplint.Edits.applyEditList, batch loop w/ parse-regression safety valve — never invents rewrites) · released.ts (released-API lookup over the bundled snapshot) · scaffold.ts (RAP templates + self-validation) · unittest.ts (ABAP Unit skeletons: fail-loudly markers, abaplint round-trip) · deps.ts (tiered dependency graph: AST db/func refs + outline structure + labeled textual) · rules.ts · formatter.ts · outline.ts (+ outlineToMermaid)src/data/— bundled data assets:released-apis.json(compact SAP Cloudification snapshot, Apache-2.0; auto-refreshed by the weekly abapmcp-upgrade cron before each release) +table-successors.json(curated classic-table → CDS successor map) +rewrite-recipes.json(curated per-category before/after Cloud rewrites, surfaced on readiness categories + plan items)scripts/—build-released-api-index.mjs(dev: fetch+transform SAP data) ·copy-data.mjs(build: copysrc/data/*.json→dist/data/)src/tool.ts,src/errors.ts— vendored mcp-kit patterns (attributed; keep in sync by hand)src/server.ts+src/cli.ts(entry: subcommand → CLI, bare → stdio server) +src/http.ts(optional stateless Streamable HTTP entry with auth/body/concurrency/rate limits) +src/cli-commands.ts(setup/lint/fix [--write]/readiness/plan/compare/scaffold/unittest/deps/ outline/explain/rules; the CLI may touch fs + spawn editor CLIs —setupregisters the server viacode --add-mcp/claude mcp addand prints guided Eclipse steps; the MCP server never does) +src/index.ts(library exports). Newcomer walkthrough:docs/INSTALL.md.docs/DESIGN.md— decision log ·docs/COOKBOOK.md— user recipesplugins/abap-mcp/+.agents/plugins/marketplace.json— Codex plugin, 4 skills (mirroring the prompts), local MCPevals/routing/— offline metadata-routing fixtures and reportsexamples/claude-code/— subagents (reviewer, migrator), .mcp.json, CI workflow
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.
- 8d ago First seen · 92 lines · 1,934 tokens per session scan A fb04d6420c9d
abap-mcp AGENTS.md is an instructions file published in the GitHub repository palimkarakshay/abap-mcp (2 stars, last pushed 15d ago), licensed MIT. It adds 1,934 tokens to every session, about $0.0097 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 2026-08-31.
Other instructions, from other repositories
next.js AGENTS.md
AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
vscode buildNext.instructions.md
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).
vscode oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).
langchain AGENTS.md
AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.
deepseek-harness AGENTS.md
AGENTS.md instructions for deepseek-ai/deepseek-harness, covering agents.md, pre-stable apis and released session data, repository layout, commands and host sandbox failures.