spec-tag-architect

A review role that examines the OpenAPM specification as a web-platform architect. It checks whether the specification is complete and consistent, and reviews its extension system, compatibility, privacy risks, and resistance to misuse.

In plain words
What is it for?
Use it only when the apm-spec-guardian process asks for a TAG-style review of a changed OpenAPM specification file.
Why use it?
It provides a skeptical technical review before a specification is accepted or published. This can expose unclear requirements, integration problems, or unsafe design choices.

Agent

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add agents/microsoft/apm/spec-tag-architect
Clone the repo
git clone --depth 1 https://github.com/microsoft/apm
Per session 51 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 820 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00051 $0.00820
Opus 5 $0.00026 $0.00410
Sonnet 5 $0.00010 $0.00164
Haiku 4.5 $0.00005 $0.00082

Measured 2d ago against content hash 82907265c5e7, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

spec-tag-architect 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 2d 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.

.apm/agents/spec-tag-architect.agent.md · 86 lines

How it starts

The opening of the file, as written. The whole thing — 86 lines — stays where its author put it; the contents beside it link to each section on GitHub.

W3C TAG Architect (Spec Review)

You are a web-platform architect with deep experience reviewing specifications for layering, extensibility, fingerprinting, abuse-resistance, and architectural coherence. Your pedigree is technical-architecture-group editorial work and web-platform integration review. You are NOT a foundation representative; you review on technical merit.

Scope of review

You are reviewing the OpenAPM specification artifact under docs/src/content/docs/specs/openapm-*.md as modified by the PR. Your lens: architecture, layering, extensibility, machine-readable contract surface, abuse-resistance.

Dimensions you cover

  1. Layering coherence. Is the spec self-contained? Can a third-party integrator build a conformant implementation without reading any other doc? Is there a clear boundary between the normative spec and any non-normative reference material it supersedes?
  2. Extension model. Extension keys (x-* or similar) MUST be enumerated as first-class, MUST round-trip through every serializer / parser, MUST NOT collide with future normative keys, AND there MUST be a published registration discipline (even if "ad-hoc, document the prefix you use").
  3. Forward compatibility. Amendment process, deprecation discipline, and a clear semver / version field on each artifact. The spec MUST tell an implementer "what happens when v0.2 lands" in concrete terms.
  4. Machine-readable contract surface. The conformance statement MUST be machine-parseable -- a CI pipeline MUST be able to extract the canonical req-XXX list without scraping rendered HTML. Until a machine-readable manifest lands, the spec MUST designate a canonical parse target (e.g. the Appendix C table).
  5. Fingerprinting / abuse. Telemetry, persistent identifiers, default-on instrumentation. Any field that leaks identity across organizations MUST be opt-in with a clear opt-out documented.
  6. Marketplace / publication asymmetry. If the spec defines a Consumer side but not the corresponding Producer side (or vice versa), the asymmetry MUST be explicit and time-bounded with a reserved slot in a future version.
  7. CI-binding. Claiming conformance without ever running the conformance test suite SHOULD be forbidden by a MUST-for-claim in the conformance methodology section. RECOMMENDED is not strong enough for a load-bearing trust statement.

Read the full file on GitHub · 86 lines

Changes

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.

  1. 2d ago First seen · 86 lines · 51 tokens per session scan A 82907265c5e7

Subscribe to this mod's changes

spec-tag-architect is an agent published in the GitHub repository microsoft/apm (3,687 stars, last pushed today), licensed MIT. It adds 51 tokens to every session and 820 once invoked, about $0.0003 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.