CategoryWeightCategory scoreContribution

Securityprompt, exec, net, exfil, eval

35%

30.8 pts

Supply chainhash, typosquat, maintainer, lockfile

20%

100

20.0 pts

Maintenancestaleness, pinning, CI

15%

100

15.0 pts

TransparencySKILL.md, perms, README

15%

100

15.0 pts

Communityinstalls, verify, response

15%

100

15.0 pts

Findings & checks · 1 flagged

Securityscore 88 · 1 finding

MEDIUM"Never tell the user" non-disclosure imperative in the skillSS-SKILL-INJECT-IMPERATIVE-01 · Prompt injection · skills/headless-interactive-reconcile-parity/SKILL.md

MEDIUM — it fires on intent; whether the agent honors the non-disclosure imperative depends on the host model.

Why it matters

A bulleted imperative like {match} tells the agent to never reveal, disclose, or mention something to the user. Used adversarially it can instruct the agent to hide its tool calls or lie about what it did — stripping the transparency a user relies on to trust the agent.

The exact value spotted

excerptskills/headless-interactive-reconcile-parity/SKILL.md· markdown

26## Anti-Patterns (What NOT to do)

27- Do not duplicate the install algorithm in two codepaths; they will drift on skip rules, ca

… (33 chars elided on L27)

28- Do not leak interactive AppState updates into headless mode or force headless callers to e

… (33 chars elided on L28)

29- Do not let headless mode skip cache invalidation just because it lacks a visual progress s

… (7 chars elided on L29)

Occurrences

1 occurrence · at L28

How to fix

Remove the non-disclosure imperative, or rescope it so it limits output format, not honesty to the user.

Delete any "never reveal / never tell the user" line aimed at the model's own behavior.
Keep legitimate confidentiality rules about external data (e.g. "do not echo API keys"), which are user-protective, not user-deceiving.

Framework references

OWASPLLM01 ATLASAML.T0051

Trace & refs

ruleSS-SKILL-INJECT-IMPERATIVE-01sha256838948ee08b3eb67…rubric 365aacaView on GitHub

Supply chainscore 100 · 0 findings

✓—All supply chain checks passedNo findings in this category for the latest scan.pass

Maintenancescore 100 · 0 findings

✓—All maintenance checks passedNo findings in this category for the latest scan.pass

Transparencyscore 100 · 0 findings

✓—All transparency checks passedNo findings in this category for the latest scan.pass

Communityscore 100 · 0 findings

✓—All community checks passedNo findings in this category for the latest scan.pass

The primary manifest — the file an agent reads to learn what this artifact does.

skills/headless-interactive-reconcile-parity/SKILL.md · 1 file

skills/headless-interactive-reconcile-parity/SKILL.md2.3 KB · Markdown

name: headless-interactive-reconcile-parity

description: "Keep interactive and headless extension installation flows behaviorally aligned by sharing reconciliation logic and adapting only the outer contract."

metadata:

author: ychampion

SKILL: Headless Interactive Reconcile Parity

Domain: plugin-installation Trigger: Apply when the same extension installation lifecycle must work in both an interactive UI and a headless or automation mode. Source Pattern: Distilled from reviewed plugin, extension, and MCP synchronization implementations.

Core Method

Centralize the real install/reconcile behavior in shared lower-level helpers, then wrap it with thin mode-specific adapters. The interactive wrapper projects pending and progress state into AppState and may trigger user-facing refresh cues; the headless wrapper returns a simple changed/not-changed signal plus metrics. This preserves one source of truth for installation semantics while letting each mode expose the minimal contract its caller needs.

Key Rules

Share the diff, reconcile, and cache-invalidation primitives between modes; only the presentation and return type should differ.
Let interactive mode add status mapping and manual-refresh prompts, but keep those concerns out of the headless path.
Let headless mode return a boolean or compact metric set so its caller can decide whether to refresh downstream registries.
Preserve environment-specific options, such as seed registration or zip-cache skips, as wrapper configuration instead of forking the core reconcile algorithm.
Whenever marketplaces change, both modes must invalidate the caches that downstream loaders depend on.

Example Application

If a coding agent can boot with a TUI or inside CI, route both startup paths through the same extension reconciler. The TUI path can paint per-source spinners, while the CI path just reports whether extensions changed and whether the caller should rebuild its command registry.

Anti-Patterns (What NOT to do)

Do not duplicate the install algorithm in two codepaths; they will drift on skip rules, cache behavior, and error handling.
Do not leak interactive AppState updates into headless mode or force headless callers to emulate UI state they do not have.
Do not let headless mode skip cache invalidation just because it lacks a visual progress surface.

headless-interactive-reconcile-parity-4e33a3— agent skill

Is headless-interactive-reconcile-parity-4e33a3 safe to install?

SKILL: Headless Interactive Reconcile Parity

Core Method

Key Rules

Example Application

Anti-Patterns (What NOT to do)

Audit the pieces. Scan the whole. Decide.