← Plugin catalog
Developer Tools

CompText Evidence

CompText Labs v0.1.0

Follow this plugin to keep it in your list and review observed changes. Free account · Up to 10 plugins · No card required.

What changed

Observed changes, with before-and-after evidence
CompText Evidence

Oct 9, 2026 · 2 saved observations

Metadata · Oct 9, 2026 · 00:06 UTC · Compared with Sep 30, 2026 · 23:14 UTC

WHAT CHANGED · RULE-BASED ANALYSIS

Package or technical metadata updated

Discoverability changed from “UNLISTED” to “LISTED”.

Observed in package metadata. These changes alone do not establish a new customer-facing feature.

Short excerpts from up to four changed fields. Open the full comparison for all changes.

Discoverability

Before

Sep 30, 2026 · 23:14 UTC

UNLISTED

After

Oct 9, 2026 · 00:06 UTC

LISTED

Share url

Before

Sep 30, 2026 · 23:14 UTC

Not present

After

Oct 9, 2026 · 00:06 UTC

https://chatgpt.com/plugins/plugins_6a8b1cb4ad708191b172159923fa5f5b?open_in_app

Metadata: full comparison →Listing: full comparison →

Publisher description

From the marketplace listing

Deterministically check whether verification evidence is newer than the latest successful mutation before accepting a completion claim.

Language: English · Automatically detected from descriptions.

Publisher keywords

Search terms declared by the publisher.

Matches for “ai”

Exact text from the indicated source. A mention alone does not establish support for your task.

Publisher description

Verify evidence freshness before accepting AI completion claims.

Publisher full description

Deterministically check whether verification evidence is newer than the latest successful mutation before accepting a completion claim.

Files & skills

File archives

Plugin package15 files · 3.3 KBBrowse files →
Skill instructions
verify-freshness1.24 KB

View saved version →

---
name: verify-freshness
description: Use when the user explicitly asks whether verification evidence is fresh enough to support an AI completion claim after workspace mutations.
---

# Verify evidence freshness

For bundled fixtures, run exactly:
`node "${CODEX_HOME:-$HOME/.codex}/plugins/cache/comptext-marketplace/comptext-evidence/0.1.0/scripts/freshness-gate.mjs" "${CODEX_HOME:-$HOME/.codex}/plugins/cache/comptext-marketplace/comptext-evidence/0.1.0/fixtures/<fixture>.jsonl"`
Use only `stale-after-mutation`, `fresh-after-verification`, or `failed-mutation-does-not-stale`; do not inspect bundled source unless the command fails.

- For user traces, require bounded content-free JSONL containing only sequence, event type, and success state.
- A successful mutation makes prior verification stale; a failed mutation does not.
- Only successful verification after the latest successful mutation clears pending state; assertion never clears it.
- Do not persist prompts, source text, raw commands/output, credentials, or environment values in traces.
- Return trace digest, mutation/verification sequence, pending state, completion verdict, and reason.
- For Riqor comparisons, report compatibility or divergence; never claim this implementation is Riqor.

Referenced files: 1

Package details

Publisher declarations from the archived package. These are separate from our research and the live service's terms.

Package author
CompText Labs
Keywords
See publisher keywords

Package observed Oct 10, 2026.

Technical details
First seen
Sep 30, 2026 · 22:02 UTC
Last seen
Oct 10, 2026 · 18:00 UTC
Collection status
Collected

plugins_6a8b1cb4ad708191b172159923fa5f5b

Download plugin data (JSON)

Before you connect CompText Evidence

How do I connect it?

Open the publisher's marketplace listing to check current availability and follow its connection instructions. This directory does not install plugins. Check the requested access and any account requirements before connecting.

Check marketplace availability ↗

Does it require paid access?

We have not established the pricing or subscription requirements for this plugin. An absent price does not mean free access.

Compare researched pricing and access models →

How can I evaluate it?

Check the declared skills and available files, then try a small task whose result you can verify. Our archived descriptions and instructions establish publisher claims, not tested runtime quality. Review sources and coverage limits.