Skip to content
Report library
Purpose / Other

Higgsfield Soul Id Skill Security Audit

What the author says it does (original text)

|

Independent security check

Do not install or run it yet

Files checked
3
Risks found
3
Could it run dangerous commands?Looks for programs run straight after downloading, remote control of your computer, and hidden commands.Risks found: 2
High risk

Runs a downloaded program right away

Source references: 2
What we found

This line runs the downloaded content without giving you a chance to read it first.

Why this matters

Once run, the downloaded program could read or change your files. The owner of the website can also replace the program without changing its URL.

This is an active installation step: when `higgsfield` is absent from PATH, the Skill pipes a script returned from GitHub's `main` branch directly into a shell. It is not pinned to a version or checksum; if the remote script or distribution chain changes, it runs with the current user's permissions and could access or modify reachable files. The user can ask for a pinned release and checksum, or download and inspect the script before running it.

SKILL.md:29In the instructionsOpen original file
   ```bash   curl -fsSL https://raw.githubusercontent.com/higgsfield-ai/cli/main/install.sh | sh   ```
Show 1 other places
SKILL.md:27In the instructionsOpen original file
1. If `higgsfield` is not on `$PATH`, install it:   ```bash   curl -fsSL https://raw.githubusercontent.com/higgsfield-ai/cli/main/install.sh | sh   ```2. If `higgsfield account status` fails with `Session expired` / `Not authenticated`, ask the user to run `higgsfield auth login` (interactive) and wait for confirmation.
High risk

Downloads a mutable remote script and executes it directly in a shell

Source references: 1
What we found

When `higgsfield` is missing, the Skill downloads an installer from the GitHub `main` branch and immediately pipes it to a shell. The content is neither pinned to a version nor verified by checksum, and the user cannot review the retrieved script before execution.

Why this matters

If the repository, release process, network connection, or installer is compromised, the script could run arbitrary commands with the user's permissions, access readable files, make persistent changes, or steal credentials.

The candidate accurately describes conditional remote execution: if the CLI is missing, the Skill fetches an installer from the mutable `main` branch and immediately runs it through `sh`. The shown command has no version pin, hash verification, or review step, creating a supply-chain risk if upstream content changes and then executes with the user's permissions. The user can restrict installation to a pinned release with a verified checksum.

SKILL.md:25In the instructionsOpen original file
Before any other command:1. If `higgsfield` is not on `$PATH`, install it:   ```bash   curl -fsSL https://raw.githubusercontent.com/higgsfield-ai/cli/main/install.sh | sh   ```2. If `higgsfield account status` fails with `Session expired` / `Not authenticated`, ask the user to run `higgsfield auth login` (interactive) and wait for confirmation.
Could it expose your files or keys?Looks for uploads of files containing passwords or keys, and keys written directly in the code.Risks found: 1
High risk

Face photos are uploaded to create a reusable identity model

Source references: 6
What we found

The Skill collects 5–20 clear, single-person face photos from varied angles, and the CLI automatically uploads any local paths. It does not require confirmation of the depicted person's consent or state retention, deletion, or provider reuse terms before upload.

Why this matters

The photos and trained identity model can support ongoing generation of identity-faithful content. For either the user or another depicted person, this may expose sensitive identity data and enable non-consensual impersonation, deepfakes, or long-term retention.

The Skill's core behavior is to train a reusable identity model from 5–20 face photos, and it explicitly says local paths are automatically uploaded by the CLI. This sends sensitive biometric material to an external service. The supplied instructions do not show a subject-consent check, retention period, deletion process, or limits on provider use, so the user should confirm these terms with the author or provider before submission and use only their own or explicitly authorized photos.

SKILL.md:21In the instructionsOpen original file
Train a face-faithful identity model. Reusable across all Soul-powered generations.
Show 5 other places
SKILL.md:43In the instructionsOpen original file
1. **Get name.** One word, used for later reference. Ask if missing.2. **Get photos.** 5–20 face photos, varied angles and lighting. Local paths or already-uploaded IDs both work — `--image` accepts either.3. **Pick variant.**
SKILL.md:49In the instructionsOpen original file
   Choose based on user's stated downstream use. Default to `--soul-2`.4. **Submit.**   ```bash   higgsfield soul-id create --name "<name>" --soul-2 --image ./photo1.png --image ./photo2.png ...   higgsfield soul-id create --name "<name>" --soul-2 --image <upload_id> --image <upload_id> ...   ```   CLI auto-uploads paths. Captures returned reference id.5. **Wait.** `higgsfield soul-id wait <id>`. Silent. Default timeout 30m.
references/photo-guide.md:10In the instructionsOpen original file
## Content- Clear face, eyes visible.- Single person per photo.- No heavy filters, no sunglasses.
SKILL.md:44In the instructionsOpen original file
1. **Get name.** One word, used for later reference. Ask if missing.2. **Get photos.** 5–20 face photos, varied angles and lighting. Local paths or already-uploaded IDs both work — `--image` accepts either.3. **Pick variant.**
SKILL.md:51In the instructionsOpen original file
   ```bash   higgsfield soul-id create --name "<name>" --soul-2 --image ./photo1.png --image ./photo2.png ...   higgsfield soul-id create --name "<name>" --soul-2 --image <upload_id> --image <upload_id> ...   ```   CLI auto-uploads paths. Captures returned reference id.5. **Wait.** `higgsfield soul-id wait <id>`. Silent. Default timeout 30m.
Could it delete files or keep running?Looks for broad file deletion, disk overwrites, and programs set to start automatically.No risks found
Could it bypass safety checks?Looks for skipped website security checks, excessive file access, or actions that skip your approval.No risks found
Could it mislead the AI or hide text?Checks the skill instructions for requests to ignore you, influence the report, or hide text in invisible characters.No risks found
Could it change links or payment recipients without asking?Looks for forced referral or payment changes combined with instructions to hide the change.No risks found

Inside this skill

7 instruction sections

The Skill submits 5–20 face photos to Higgsfield to train an identity model that can be reused in later image or video generation.

View source
SKILL.md:5In the instructionsOpen original file
description: |  Train a Soul Character — a personalized model on a person's face that  Higgsfield uses for identity-faithful image and video generation.  Use when: "create my Soul", "train my face", "make my digital twin",  "build me an avatar", "learn my appearance", "create a character of me",  "set up identity for video", "I want my face in generated images".  Chain: train Soul (one-time, returns reference_id) → use in  higgsfield-generate via `--soul-id <id>` with models like  `text2image_soul_v2` or `soul_cinema_studio`.  NOT for: one-shot face swaps (use higgsfield-generate with --image),
SKILL.md:43In the instructionsOpen original file
1. **Get name.** One word, used for later reference. Ask if missing.2. **Get photos.** 5–20 face photos, varied angles and lighting. Local paths or already-uploaded IDs both work — `--image` accepts either.3. **Pick variant.**   - `--soul-2` — for image generation (default)   - `--soul-cinematic` — for cinematic / video work   Choose based on user's stated downstream use. Default to `--soul-2`.4. **Submit.**

When local paths are submitted, the CLI automatically uploads the photos, returns an identity-model reference ID, and then waits for training for about 30 minutes by default.

View source
SKILL.md:49In the instructionsOpen original file
   Choose based on user's stated downstream use. Default to `--soul-2`.4. **Submit.**   ```bash   higgsfield soul-id create --name "<name>" --soul-2 --image ./photo1.png --image ./photo2.png ...   higgsfield soul-id create --name "<name>" --soul-2 --image <upload_id> --image <upload_id> ...   ```   CLI auto-uploads paths. Captures returned reference id.5. **Wait.** `higgsfield soul-id wait <id>`. Silent. Default timeout 30m.6. **Deliver.** "Soul `<name>` ready. Use in generate with `--soul-id <id>`."

The service requires authentication and a Basic-or-higher paid plan for training; the Skill says to inform free-plan users before submission.

View source
SKILL.md:31In the instructionsOpen original file
   ```2. If `higgsfield account status` fails with `Session expired` / `Not authenticated`, ask the user to run `higgsfield auth login` (interactive) and wait for confirmation.3. Soul training requires a paid plan (Basic+). If `higgsfield account status` shows free plan, tell the user before submitting.
Start here · InstructionsSKILL.md
higgsfield-soul-id
Lines connect the instruction file to its sections, not an observed execution order. Select a section to read the source.

File reference map

References: 2
Files making referencesReferenced content
Lines show actual file references, not execution order. Select a node to highlight its connections and inspect the files and source locations. Dashed lines include files that still need locating.
Files and check records3 files

Coverage and gaps

Content covered in each file

These are the source ranges included in this check, not a guarantee that every issue has been resolved.

  • SKILL.mdFull text included
  • references/photo-guide.mdFull text included
  • references/troubleshooting.mdFull text included

This report is for the version above. We read the available code and instructions without running the skill or checking extra packages it installs. This is not a promise of safety: a different version or setup may behave differently.

  • SKILL.mdInstructions
  • references/photo-guide.mdSupporting file
  • references/troubleshooting.mdSupporting file

Operations mentioned in code and instructions

Run commands
SKILL.md:16In the instructionsOpen original file
argument-hint: "[name] [photo paths...]"allowed-tools: Bash---
SKILL.md:28In the instructionsOpen original file
1. If `higgsfield` is not on `$PATH`, install it:   ```bash   curl -fsSL https://raw.githubusercontent.com/higgsfield-ai/cli/main/install.sh | sh
SKILL.md:50In the instructionsOpen original file
4. **Submit.**   ```bash   higgsfield soul-id create --name "<name>" --soul-2 --image ./photo1.png --image ./photo2.png ...
Connect to websites
SKILL.md:29In the instructionsOpen original file
   ```bash   curl -fsSL https://raw.githubusercontent.com/higgsfield-ai/cli/main/install.sh | sh   ```
Lines read
147
File checksum (to compare versions)
6cd54a5674932bc748b9a3bc73e897e718d14c3528af223185ca00bfdd7712c6