Powered by OpenAIRE graph
Found an issue? Give us feedback
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/ ZENODOarrow_drop_down
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
ZENODO
Other literature type . 2026
License: CC BY
Data sources: ZENODO
ZENODO
Other literature type . 2026
License: CC BY
Data sources: Datacite
ZENODO
Other literature type . 2026
License: CC BY
Data sources: Datacite
versions View all 2 versions
addClaim

KCS-ALP-L1: Agent Lockdown Profile (ALP-L1) v1.0.0 — Public Draft (Unsigned Preview)

Fail‑closed agent runtime lockdown — deterministic receipts + buyer‑run offline verification.
Authors: Lee, Yong Bok;

KCS-ALP-L1: Agent Lockdown Profile (ALP-L1) v1.0.0 — Public Draft (Unsigned Preview)

Abstract

Public Draft (Unsigned Preview) — Evaluation-only.For procurement/conformance: treat as UNTRUSTED → EXPECT HOLD (fail-closed).NOT FOR INCORPORATION BY REFERENCE. Procurement-grade status applies only to a future signed release with verifiable release-gate authenticity artifacts. Receipts, not promises. What this is KCS-ALP-L1: Agent Lockdown Profile (ALP-L1) v1.0.0 (Public Draft) — a procurement-shaped interoperability profile for enterprise agent runtimes, enabling buyer-run, offline verification of conformance receipts and replay determinism. Included artifacts: Whitepaper PDF (normative), bundle ZIP (offline verifier + fixtures), Release Checklist, publication‑integrity addendum, artifact‑signing public key, and CITATION.cff. Release date: 2026‑02‑18 Canonical standards surface: https://meridianverity.com/standards/ Project home: https://meridianverity.com/ Pinned artifact‑signing fingerprint (trust anchor): 9ACD3C8B2E02BD4CAA6742EB7132DE1CAA7BBB01 Normative scope Only the Whitepaper PDF is normative. All other files in this record are supporting artifacts (bundle ZIP, offline verifier + fixtures, templates, addenda, and reviewer/auditor packets). Fail‑closed posture: HOLD blocks sensitive side effects by default when evidence is missing, stale, inconsistent, unverifiable, or non‑deterministic. Fail‑closed / release gate This deposit is UNSIGNED_PREVIEW. Consumers MUST treat this deposit as UNTRUSTED → HOLD (fail‑closed) for any conformance, safety, or procurement acceptance claim. For a future signed release, procurement‑grade status exists if and only if offline verification succeeds under the pinned fingerprint: sha256sum -c SHA256SUMS gpg --verify SHA256SUMS.asc SHA256SUMS What the profile specifies (high level) · Deterministic runtime verdicts (ALLOW / DENY / HOLD) with standardized reason codes. · Deterministic offline outcomes (PASS / FAIL / HOLD) for Evidence Pack verification and replay checks. · Signed allowlist‑only tool/skill invocation (tamper‑evident). · Permit‑before‑send network egress (deny‑by‑default). · Untrusted → trusted boundary enforcement for privileged actions. · Secret isolation + scoped use (no plaintext secrets in prompts, tool output, or logs). · High‑risk action approval gates (HOLD until approval proof exists). · Version pinning + drift detection; stewardship/change‑control expectations. · Portable Evidence Packs with minimal schema/registry surfaces designed for procurement attachment. Interoperability evidence (included) This record includes a reference offline verifier contract and portable fixture Evidence Packs demonstrating expected outcomes (PASS / HOLD / FAIL), including negative and boundary cases, with deterministic receipts + reason codes. Buyer‑run demo from bundle root: python3 verifier_contract/alp_l1_offline_verifier.py verifier_contract/fixtures/ALP_SAMPLE_PACK_TV-ALP-001_PASS_BASELINE_v1.0.0.zip python3 verifier_contract/alp_l1_offline_verifier.py verifier_contract/fixtures/ALP_SAMPLE_PACK_TV-ALP-002_HOLD_ALLOWLIST_SIG_MISSING_v1.0.0.zip python3 verifier_contract/alp_l1_offline_verifier.py verifier_contract/fixtures/ALP_SAMPLE_PACK_TV-ALP-005_FAIL_DIGEST_MISMATCH_v1.0.0.zip Security considerations (snapshot) Threat model coverage includes (illustrative): tool injection/capability sprawl, allowlist tampering, data exfiltration, prompt injection, secret leakage, unauthorized high‑risk actions, baseline drift/downgrade, evidence repudiation, and TOCTOU‑style drift. Public‑safe by design The profile and artifacts avoid exploit guidance and do not require disclosure of confidential implementation details by default. License / rights notice CC BY 4.0 applies to text and supporting artifacts unless a file states otherwise. No patent license by publication. Not legal advice. Integrity / verification Use SHA256SUMS to verify file integrity (SHA‑256) after download/extraction: sha256sum -c SHA256SUMS How to cite Use the Zenodo “Cite as” entry after publication. CITATION.cff is included for convenience.

Keywords

Procurement, Governance, Runtime Security, Evidence Pack, Fail-Closed, Agentic AI, AI Safety, Offline Verification, Audit, Interoperability

  • BIP!
    Impact byBIP!
    selected citations
    These citations are derived from selected sources.
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    0
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Average
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Average
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Average
Powered by OpenAIRE graph
Found an issue? Give us feedback
selected citations
These citations are derived from selected sources.
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
0
Average
Average
Average