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.
npx agentmods add skills/mystilleef/spae-framework/spae-verifynpx skills add mystilleef/spae-framework --skill spae-verifygit clone --depth 1 https://github.com/mystilleef/spae-frameworkWrote 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/skills/mystilleef/spae-framework/spae-verify)<a href="https://agentmods.dev/skills/mystilleef/spae-framework/spae-verify"><img src="https://agentmods.dev/badge/skills/mystilleef/spae-framework/spae-verify.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 | $0.00026 | $0.01978 |
| Opus 5 | $0.00013 | $0.00989 |
| Sonnet 5 | $0.00005 | $0.00396 |
| Haiku 4.5 | $0.00003 | $0.00198 |
Grade C, and why
spae-verify scanned grade C with 1 finding 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 5d 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.
Hidden instructionshighPrompt injection
Directives inside HTML comments, invisible characters or bidirectional overrides are read by the model and not by the person reviewing the file.
<!-- prettier-ignore-start --> How it starts
The opening of the file, as written. The whole thing — 190 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Verify SPAE
When to use
STATE.jsonreportsphase: verify.- Every
PLAN.mdtask reportsdone.
Goal
- Compare implemented repository state against
SPEC.md, identify gaps, and close or reopen the work stream as the finalSPAEarbiter. - Pass: complete the work stream.
- No pass: route actionable gaps back to
/spec.
Input
Read:
.spae/current/STATE.json(consultreferences/STATE.mdfor field reference, directives, and phase snapshots).spae/current/SPEC.md.spae/current/PLAN.md- All source files modified by this workstream (scoped from
.spae/current/PLAN.mdtasks) .spae/current/VERIFY.md, when present
Workflow
- GATE—Read
.spae/current/STATE.json; confirmphase: verify. Confirm every.spae/current/PLAN.mdtask reportsdonein thetasksregistry. Halt immediately on failure; report which check failed; make no changes. - ORIENT—Goal: compare implemented repository state against each
.spae/current/SPEC.mdspec item; determine pass, no-pass, or blocked verdict. - PLAN—List
SPEC.mdspec item IDs and source files modified by thisworkstream(scoped fromPLAN.mdtasks). - ACT—Execute:
- Run all relevant project checks. When no automated checks exist,
inspect observable behavior directly—reading code and logs or
running scripted probes—never delegating execution or evidence
capture to a human, an attended terminal, or an interactive
session. Record all results; any check failure mapping to a
SPEC.mditem classifies as a hard block. - When a
SPEC.mditem resists verification without human execution, an attended or interactive terminal, or human presence: classify it as a hard block and record it inVERIFY.mdas a spec defect for/specto resolve; never instruct the user to perform a verification step. - Compare implementation against each
SPEC.mditem using check results as evidence. Reference each finding by spec item ID. - When
.spae/current/VERIFY.mdexists: treat each prior finding as an explicit re-check item; confirm each addressed before proceeding. - Classify each finding:
- Hard block: regression, contract break, missing or incorrect
required behavior, check failure mapping to a
SPEC.mditem; drives verdict to no-pass. - Soft finding: absent test coverage of explicit required
behavior; unsafe optimization; complexity ungrounded in an
explicit
SPEC.mdspec item (for example, a validation library pulled in for a single basic check); note only; no verdict impact. - Observation: thin edge-case test coverage, ambiguous or
untestable spec item, minor deviation outside SPEC scope,
unrequestedguardrail, theoretical risk; note only; no verdict impact.
- Hard block: regression, contract break, missing or incorrect
required behavior, check failure mapping to a
- Draft findings against the
references/VERIFY.mdschema; consultreferences/prose-protocol.mdfor phrasing.
- Run all relevant project checks. When no automated checks exist,
inspect observable behavior directly—reading code and logs or
running scripted probes—never delegating execution or evidence
capture to a human, an attended terminal, or an interactive
session. Record all results; any check failure mapping to a
- VERIFY—Confirm any drafted
VERIFY.mdcontent matchesreferences/VERIFY.mdstructure. - PERSIST—Apply verdict:
- Pass: set
STATE.jsontostatus: completed,phase: done; remove.spae/current/VERIFY.mdwhen present; remove.spae/currentsymlink. - No pass: create or overwrite
.spae/current/VERIFY.mdusing thereferences/VERIFY.mdschema with hard blocks from the current run only, followed by soft findings and observations as informational notes; setSTATE.jsontostatus: revision_required,phase: spec,cursor: {},tasks: {}. - Blocked: write blocker details only to
.spae/current/VERIFY.md(omit observations); setSTATE.jsontostatus: revision_required,phase: spec,cursor: {},tasks: {}. - All verdicts: re-read
STATE.jsonafter writing; on a field mismatch, rewrite and re-read. On Pass, confirm the re-read before removing.spae/current.
- Pass: set
- REPORT—Emit the result following the result directives and using the result template. On pass, surface observations under Findings. On blocked, emit the Blocked result block.
What ships with it
4 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 5d ago First seen · 190 lines · 26 tokens per session scan C 44529e3dc89d
spae-verify is a skill published in the GitHub repository mystilleef/spae-framework (1 stars, last pushed 1mo ago), licensed MIT. It adds 26 tokens to every session and 1,978 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it C with 1 finding (hidden instructions). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
architecture
Project architecture and file structure conventions for all process types. Use when: (1) Creating new files or modules, (2) Deciding where code should go, (3) Converting single-file components to directories, (4) Reviewing code for structure compliance, (5) Adding new bridges, services, agents, or workers.
testing
Testing workflow and quality standards for writing and running tests. Use when: (1) Writing new tests, (2) Adding a new feature that needs tests, (3) Modifying logic that has existing tests, (4) Before claiming a task is complete.
ahk-test
Design, write, and run behavior-focused tests for an objective or existing code. Writes test files only and reports evidence. No tasks created, no harness tracking.
ahk-review
Preview a code review against ticket/objective alignment, with deep semantic (name-vs-behavior) analysis. No tasks created, no harness tracking.
ahk-use-cases
Define or refine a feature, change, refactor, or user journey as an approved non-technical use-case specification.
ahk-triage
Triage a bug or unexpected behavior. Deep diagnostic analysis with structured report. No tasks created, no harness tracking.