Agento-patronum patronum-add

Add a file pattern or command to the agento-patronum protection list (blacklist) or allowed list (whitelist). Use when the user wants to block or explicitly allow access to a file, path, or command.

install

source · Clone the upstream repo

git clone https://github.com/emaarco/agento-patronum

Claude Code · Install into ~/.claude/skills/

T=$(mktemp -d) && git clone --depth=1 https://github.com/emaarco/agento-patronum "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/patronum-add" ~/.claude/skills/emaarco-agento-patronum-patronum-add && rm -rf "$T"

manifest: skills/patronum-add/SKILL.md

source content

Skill: patronum-add

Add a pattern to the agento-patronum blacklist (block) or whitelist (allow).

Steps

1. Parse input

Parse the user's input from

$ARGUMENTS

. Determine two things: the pattern and the target list.

Target list detection

If the user's intent uses words like "allow", "whitelist", "permit", "trust", "safe to access", or "let Claude read", they want the whitelist. Add

--whitelist

to the command.

If the intent uses words like "protect", "block", "hide", "deny", "prevent", or "restrict", they want the blacklist (default).

Pattern detection: two input modes

Mode A — Exact glob or path (e.g.

**/*.tfstate

~/.aws/credentials

): Use it directly as the pattern. If no

--reason

is provided, generate a short one.

Mode B — Natural language intent (e.g.

I want to protect my terraform state files

): Detect that the input is not a glob/path (no

, or

.**

characters, or it reads as a sentence). Derive one or more appropriate glob patterns from the intent. Common mappings:

"terraform state" →
```
**/*.tfstate
```
,
```
**/*.tfstate.backup
```
"AWS credentials / secrets" →
```
~/.aws/credentials
```
,
```
~/.aws/config
```
"environment variables / .env" →
```
**/.env
```
,
```
**/.env.*
```
"private keys / SSH" →
```
**/*.pem
```
,
```
**/*.key
```
,
```
~/.ssh/*
```
"docker credentials" →
```
~/.docker/config.json
```
For anything not in the list above, reason from the technology/domain to derive a sensible glob.

2. Confirm with user

Always confirm via

AskUserQuestion

— never skip this step regardless of input mode. Present:

The pattern(s) to be added
Whether it will go to the blacklist (block) or whitelist (allow)
The reason (provided or generated)
A warning if the pattern looks overly broad (e.g.
```
*
```
,
```
**/*
```
, or very short globs)

3. Add the pattern

After the user confirms, run:

node "${CLAUDE_PLUGIN_ROOT}/scripts/management/patronum-add.js" $ARGUMENTS

Include

--whitelist

in the arguments if the user wants to allow access.

4. Present result

Run

node "${CLAUDE_PLUGIN_ROOT}/scripts/management/patronum-list.js"

and present the updated protection list, showing both the blacklist and whitelist sections. Highlight the newly added entry.