Back to skill

Security audit

Tmp Soul Evolver

Security checks for vulnerabilities and agentic risk

Overview

This skill has a coherent memory-evolution purpose, but it grants unattended authority to send private memory data to an external API and persist model-generated changes into agent control files without enough safeguards.

Review this skill carefully before installing. Use dry-run only, avoid enabling the scheduled hook, restrict the workspace and API environment, and do not run it on memory files that may contain secrets or confidential conversations. It should not automatically write to SOUL.md, AGENTS.md, or similar control files until target allowlisting, redaction, endpoint validation, and explicit approval for each proposed update are added.

Vulnerability Patterns
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • System PersistenceInstalls backdoors, hooks, services, or scheduled tasks that survive the run
  • Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Findings (5)

T02 · Agent Memory Poisoning

Error
Location
soulforge/analyzer.py:278
Finding

Untrusted memory and remote-model output can poison persistent Agent instructions

Content
View full analysis
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Error
Location
soulforge/evolver.py:93
Finding

Remote-model-controlled target path permits writes outside the workspace

Content
View full analysis
Dict[str, Any]: file_path = self.workspace / filename result = {"applied": 0, "skipped": 0, "error": None} # Load existing content if file_path.exists(): existing_content = file_path.read_text(encoding="utf-8") else: existing_content = "" # Create empty file with basic structure file_path.parent.mkdir(parents=True, exist_ok=True) logger.info(f"File doesn't exist, will create: {filename}") # Check for duplicates patterns_to_apply = self._filter_duplicates(patterns, existing_content) if not patterns_to_apply: result["skipped"] = len(patterns) logger.info(f"All patterns already exist in {filename}, skipping") return result # Create backup if not dry_run and self.config.get("backup_enabled", True): self._create_backup(file_path) # Apply each pattern for pattern in patterns_to_apply: ...[truncated 2197 chars]
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Error
Location
soulforge/analyzer.py:194
Finding

Sensitive memory and identity-file content is transmitted to an external API without redaction

Content
View full analysis
str: """Prepare memory entries as text for the prompt.""" lines = [] for i, entry in enumerate(entries[:100]): # Limit to 100 entries lines.append(f"[{i+1}] Source: {entry.source} | Type: {entry.source_type} | Category: {entry.category}") lines.append(f" Content: {entry.content[:300]}") if entry.timestamp: lines.append(f" Time: {entry.timestamp}") lines.append("") return "\n".join(lines) def _format_existing_content(self, existing_content: Dict[str, str]) -> str: """Format existing content for the prompt.""" if not existing_content: return "No existing content found." lines = [] for filename, content in existing_content.items(): lines.append(f"\n### {filename}") # Show first 1500 chars of each file lines.append(content[:1500]) if len(content) > 1500: lines.append("...(truncated)...") return "\n".join(lines) ``` ```python # soulforge/analyzer.py:225-252 url = f"{self._base_url}/text/chatcompletion_v2" payload = { "model": self._model, "messages": [ {"role": "system", "content": system_prompt}, ...[truncated 1977 chars]
Remediation
View remediation

T09 · Insecure Skill Coding Practices

Error
Location
soulforge/config.py:146
Finding

Configurable API origin can disclose the bearer key and workspace data to an attacker-controlled server

Content
View full analysis
str: """Get MiniMax API base URL.""" return self._config.get( "minimax_base_url", os.environ.get("MINIMAX_BASE_URL", "https://api.minimax.chat/v1") ) ``` ```python # soulforge/analyzer.py:225-252 url = f"{self._base_url}/text/chatcompletion_v2" payload = { "model": self._model, "messages": [ {"role": "system", "content": system_prompt}, {"role": "user", "content": user_prompt} ], "temperature": 0.3, "max_tokens": 4096, } data = json.dumps(payload).encode("utf-8") headers = { "Authorization": f"Bearer {self._api_key}", "Content-Type": "application/json", } req = urllib.request.Request(url, data=data, headers=headers, method="POST") try: with urllib.request.urlopen(req, timeout=60) as resp: response_data = json.loads(resp.read().decode("utf-8")) return response_data.get("choices", [{}])[0].get("message", {}).get("content", "") ``` ### Technical Analysis `minimax_base_url` can be supplied by a configuration file, while `MINIMAX_BASE_URL` can be supplied through the process environment. The value is not restricted to the official MiniMax HTTPS origin and is used to construct a request carrying both the bearer API key and sensitive workspace data. Consequently, a malicious configuration or compromised execution environment can redirect requests to an attacker-controlled HTTP or HTTPS server. The returned JSON is then processed as trusted model output, allowing this issue to be chained with persistent memory poisoning and path traversal. ### Attack Path 1. An attacker modifies a SoulForge configuration file, supplies a malicious `--config` file, or contr ...[truncated 967 chars]
Remediation
View remediation

T06 · System Persistence

Error
Location
SKILL.md:13
Finding

Default scheduled hook repeatedly performs unattended external analysis and persistent file modification

Content
View full analysis
int: """Set up cron schedule.""" if args.every: print(f"To schedule SoulForge every {args.every} minutes, add this to your crontab:") print("") print(f"*/{args.every} * * * * cd {os.getcwd()} && python3 {__file__} run >> /var/log/soulforge.log 2>&1") print("") print("Or with OpenClaw cron:") print(f" openclaw cron add --name soulforge-evolve --every {args.every}m \\") print(f" --message 'exec python3 {__file__} run'") return 0 ``` ### Technical Analysis The Skill metadata defines a recurring hook every 120 minutes, and the documentation recommends an OpenClaw cron task. This behavior is declared and related to the stated continuous-evolution feature, but it causes sensitive processing and persistent Agent-file modification to continue across sessions without requiring review for each run. The flagged `cmd_cron()` implementation does not directly edit crontab or invoke a scheduler; it only prints commands for the user to install. The persistence concern primarily arises from the scheduled hook declared in `SKILL.md` and the recommended scheduler registration. Scheduling is especially dangerous when combined with the confirmed prompt-injection, endpoint, and path-valida ...[truncated 1081 chars]
Remediation
View remediation
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Rogue AgentSelf-Modification, Session Persistence
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
Findings (31)

Tp4

High
Category
MCP Tool Poisoning
Confidence
89% confidence
Finding

The documented behavior materially diverges from the described purpose, including missing evidence of identity-file evolution and MiniMax use while mentioning other ingestion sources like hawk-bridge. Security review depends on accurate declarations; when a skill's description does not match real behavior, operators cannot reliably assess data flows, side effects, or exposure to external systems.

Content

No source excerpt is available for this finding.

Missing User Warnings

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill description fails to clearly warn that memory contents may be transmitted to an external MiniMax API for analysis. That omission is dangerous because memory and identity files often contain sensitive user preferences, internal workflows, and potentially secrets; sending them off-box without prominent disclosure undermines informed consent and can create confidentiality and compliance issues.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The README explicitly promotes continuous analysis of memory files and use of the MiniMax API, but it does not clearly warn that potentially sensitive conversation logs, preferences, and learning records may be transmitted to a third-party service. In a memory-evolution skill, those sources are likely to contain personal, operational, or confidential data, so lack of an explicit privacy/data-sharing warning materially increases the risk of unintended disclosure.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The README advertises automatic updates to core workspace identity and memory files and later describes scheduled execution, but it does not place a prominent upfront warning that the tool will modify multiple files over time. Because these files influence future agent behavior, silent or poorly disclosed scheduled modification can cause configuration drift, integrity issues, or hard-to-trace behavioral changes in the agent workspace.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The README states that memory files are automatically analyzed using the MiniMax API, but it does not clearly warn that potentially sensitive conversation logs, learnings, or profile data may be transmitted to a third-party remote service. In a memory-evolution skill, this omission is security-relevant because users may reasonably assume local-only processing while the system persists and exports private workspace context.

Content

No source excerpt is available for this finding.

Ssd 3

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The document describes continuous analysis of memory logs, vector-memory content, and learnings to update persistent identity/profile files. This creates a real data retention and exposure risk because sensitive user information, behavioral patterns, and historical context are aggregated into durable files, increasing the blast radius of compromise or accidental disclosure.

Content

No source excerpt is available for this finding.

Ssd 3

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The update rules explicitly instruct the system to record new user preferences, projects, decisions, and work-style changes into persistent files such as USER.md and MEMORY.md. In this skill's context, that is more dangerous because the purpose of the tool is long-term identity evolution, so it normalizes accumulation of personal and operational data that may later be exposed, over-relied upon, or propagated across sessions without clear consent boundaries.

Content

No source excerpt is available for this finding.

Undeclared Tool Scope

Medium
Category
MCP Least Privilege
Confidence
88% confidence
Finding

The skill advertises executable behavior involving environment access, file reads/writes, and network use, but does not declare any explicit tool scope or allowed-tools boundary. That makes the operational blast radius unclear to users and reviewers, increasing the chance that sensitive memory files or API-backed exfiltration occur without informed consent or enforceable restrictions.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The activation conditions are broad enough to trigger during many ordinary interactions, increasing the likelihood that the skill runs without deliberate operator intent. Because this skill can read memory, modify identity files, and potentially contact an external API on a schedule, overbroad invocation materially raises the risk of unnecessary data processing and unintended persistent changes.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The documentation gives conflicting statements about agent isolation, first claiming agent-specific backup/state directories and later describing shared global paths. In a multi-agent environment, that ambiguity can lead to cross-agent data mixing, accidental overwrite of another agent's state, or disclosure of one agent's memory-derived identity data to another.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
85% confidence
Finding

The safety claim that the skill only appends and never overwrites conflicts with language stating it updates existing identity files. Misleading safety guarantees are dangerous because users may trust the tool with high-value configuration or identity files under the false assumption that destructive modification cannot occur.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The system is designed to automatically update SOUL.md, USER.md, IDENTITY.md, and similar workspace files, yet the architecture does not call out a clear warning or approval boundary for modifying user-controlled files. In the context of an identity-evolution skill, silent or poorly disclosed writes are risky because they can permanently alter behavioral instructions, preferences, or repository state in ways the user may not expect or review.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The architecture explicitly describes sending memory entries and existing identity-file content to an external MiniMax API, but it does not mention any consent, redaction, or warning to the user before transmission. Because these sources can contain sensitive conversation history, preferences, internal workflow details, or secrets embedded in workspace files, this creates a meaningful confidentiality and privacy risk in a memory-management skill whose core purpose is to aggregate such data.

Content

No source excerpt is available for this finding.

Session Persistence

Medium
Category
Rogue Agent
Confidence
60% confidence
Finding

Skill establishes unauthorized persistence across sessions via cron jobs, startup scripts, or state files. Session persistence allows an attacker to maintain access beyond the current interaction.

Content

Scanner excerpt · scripts/soulforge.py (reported line 128)May include surrounding context.

python
print("")
    if results["dry_run"]:
        print("🔍 DRY RUN complete. Run without --dry-run to write changes.")
    else:
        print("✅ SoulForge evolution complete!")

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
87% confidence
Finding

The restore command can overwrite workspace files from a user-supplied backup path without any confirmation, validation, or safety interlock. In a tool that maintains identity and behavioral files automatically, accidental or unauthorized restoration could revert important state, reintroduce poisoned content, or replace trusted files with stale or malicious backups.

Content

No source excerpt is available for this finding.

Ssd 3

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The system prompt explicitly instructs the external model to mine conversation logs, learning records, errors, and feature requests, then promote them into persistent identity files. This creates a structured natural-language exfiltration and persistence channel where sensitive user data, behavioral traits, and internal project details can be extracted, summarized, and retained long-term beyond the original context.

Content

No source excerpt is available for this finding.

Ssd 3

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The user prompt template embeds raw memory entries and existing file contents directly into a request sent to an external model, broadening disclosure to include both source memories and current identity documents. Because the prompt asks the model to compare, summarize, and generate updates from this material, it increases the chance that sensitive data is exposed, transformed, and re-persisted in additional files.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

This code sends raw memory entries and existing identity-file contents to a third-party MiniMax API, which can include sensitive user conversations, preferences, and internal workspace data. Even if this is the feature's intended behavior, the transfer happens without any explicit consent gate, redaction step, or in-code warning, so private data can be disclosed outside the local environment.

Content

No source excerpt is available for this finding.

External Transmission

Medium
Category
Data Exfiltration
Confidence
60% confidence
Finding

Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.

Content

Scanner excerpt · soulforge/config.py (reported line 150)May include surrounding context.

python
"""Get MiniMax API base URL."""
        return self._config.get(
            "minimax_base_url",
            os.environ.get("MINIMAX_BASE_URL", "https://api.minimax.chat/v1")
        )

    @property

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The class docstring explicitly lists 'Incremental updates (appends, never overwrites)' as a safety feature. However, restore_from_backup uses shutil.copy2(backup, target) to replace the target file contents, which is an overwrite operation and contradicts the stated safety guarantee.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The module docstring and class documentation say the reader 'reads memory files from multiple sources' including the hawk-bridge vector store, and read_all() implies a unified list of entries from all sources. But _read_hawk_bridge() only connects to LanceDB, counts rows, and logs the count without creating any MemoryEntry objects, so hawk-bridge content is not actually included in the returned memory set.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

This skill is designed to automatically read workspace memory logs and learning files that may contain sensitive user preferences, errors, and behavioral history, yet the code shows no consent gate, scope restriction, or disclosure mechanism before aggregation. In the context of an 'AI memory evolution' skill, silent ingestion of these files increases privacy risk because accumulated personal or operational data can be processed and potentially propagated into identity files or downstream API use.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The comments 'Try to read from LanceDB' and 'Get recent entries (last 50)' indicate that vector-store records will be read as memory content. In practice, the code merely checks table existence, computes a count, and logs it; no records are fetched into MemoryEntry objects and nothing from hawk-bridge reaches callers.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
88% confidence
Finding

The isolation table says memory sources are read only from the agent's own memory/ and .learnings/. However, the later memory sources table includes 'hawk-bridge | Vector store', which contradicts the earlier narrow description of where data comes from.

Content

No source excerpt is available for this finding.

Dynamic attribute access via getattr()

Low
Category
Dangerous Code Execution
Confidence
50% confidence
Finding

Dynamic getattr() with a non-literal attribute name can access arbitrary object attributes, potentially bypassing access controls.

Content

Scanner excerpt · scripts/soulforge.py (reported line 33)May include surrounding context.

python
def setup_logging(level: str = "INFO") -> None:
    """Configure logging."""
    logging.basicConfig(
        level=getattr(logging, level.upper(), logging.INFO),
        format="%(asctime)s [%(levelname)s] %(name)s: %(message)s",
        datefmt="%Y-%m-%d %H:%M:%S",
    )