TESP Audit

Audit whether the Task Execution Signal Protocol is still being followed, with low-token, exception-first checks for version drift, queue hygiene, numeric st...

Security Scan

This is an instruction-only governance/audit skill that appears internally consistent: it checks for specific repo files and artifacts related to TESP, asks for no installs or credentials, and writes a local report.

Purpose & Capability

The name/description (TESP audit) matches the instructions (look for protocol, registry, queue, version anchors, numeric progress, archive). There are no unexpected binaries, env vars, or installs required.

Instruction Scope

The runtime instructions explicitly direct the agent to read repository-like files (TEAM_EXECUTION_SIGNAL_PROTOCOL.md, TEAM_MODULES_REGISTRY.md, TASK_QUEUE.md, TASK_ARCHIVE.md) and to write a short local report. This is appropriate for an audit skill, but the SKILL.md is high-level and leaves the report path unspecified—so you should confirm where reports will be written and limit filesystem scope to the target workspace.

Install Mechanism

No install specification or code files are included; this is instruction-only and will not write/download artifacts during install.

Credentials

The skill requests no environment variables, credentials, or config paths. The checks it describes (file/registry/queue checks) are proportional to the stated purpose.

Persistence & Privilege

always is false and there is no special persistence requested. The skill can be invoked autonomously by the agent (platform default), which is typical for skills of this kind; this alone is not a concern.

Assessment

This skill is instruction-only and appears to only read specific repository files and write a local audit report. Before installing or enabling it broadly: 1) confirm the workspace or repo directories the agent will be allowed to read so it cannot access unrelated files; 2) test the skill on a copy of your repo to see exactly what files it reads and where it writes the report (the SKILL.md leaves the report path unspecified); 3) if you allow autonomous invocation, limit its permissions to the audit scope and periodically review generated reports; and 4) if you need stronger guarantees, ask the skill author to include explicit safe paths and an explicit report-location setting.

Use this skill when the goal is not to run a task under TESP, but to verify that TESP is still intact.

中文简介

tesp-audit 用于检查 TESP 是否仍在被正确执行，而不是直接承担任务执行本身。 它适合 rollout audit、light audit、任务板卫生检查、版本漂移识别、协议落地抽查等治理场景。 默认原则是：低 token、异常优先、常态静默。

What this skill is for

This skill checks whether the TESP system still has its critical anchors:

• protocol exists
• version is visible
• numeric progress format is intact
• active board stays clean
• completed work gets archived
• audits remain cheap by default

Default audit order

Follow this order:

1. file anchors
2. registry anchors
3. visible version anchors
4. queue hygiene
5. sampled behavior only if needed

Audit rule

Prefer:

• file checks
• diffs
• small samples

Avoid by default:

• full session replay
• large narrative summaries
• premium-model overuse for simple governance checks

Queue rule

Treat these as hygiene failures:

• completed work still sitting in active board
• missing archive destination
• malformed or non-numeric stage progress
• stale active tasks with no closure path

Model rule

Default low-cost governance models:

• GLM
• MiniMax

Escalate only when semantic conflicts or redesign decisions appear.

Read next

For the audit scope and examples, read:

• references/audit-reference.md