Back to skill

Security audit

soul-in-sapphire

Security checks for vulnerabilities and agentic risk

Overview

ClawScan marks this skill suspicious because it allows persistent edits to core context based on its own changed self-understanding, and many referenced scripts were only partially analyzed.

Consider avoiding installation until the core-context update process and the partially inspected scripts can be reviewed, especially how changes are authorized. The verdict is suspicious, not malicious, and the incomplete scan means the evidence does not establish that the skill is safe.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (36)

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 17)May include surrounding context.

md
rest_slot` v1 via `normalizeInterestSlot` / `buildContextualInterestRecall` in `scripts/memory_policy.js`), and (3) Soul ambient recall (`stage_ambient_recall.j

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 166)May include surrounding context.

md
- `memory_policy.js`: `ambientEnabled`, `normalizeInterestSlot`, and `buildContextualInterestRecall` for portable `soul_interest_slot` v1 validation and context

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 17)May include surrounding context.

md
s/memory_policy.js`), and (3) Soul ambient recall (`stage_ambient_recall.js` / `read_ambient_recall.js`, including the optional `interest` shelf on roll 7 when

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 31)May include surrounding context.

md
7. Read unexpired ambient context through `read_ambient_recall.js` without rerolling. For a Notion state candidate, use `--resolve` with events/emotions/state d

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 165)May include surrounding context.

md
- `stage_ambient_recall.js` / `read_ambient_recall.js`: ambient staging (shelves: `recent`, `unresolved`, `durable`, `dream`, and optional `interest` via `--int

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 17)May include surrounding context.

md
alInterestRecall` in `scripts/memory_policy.js`), and (3) Soul ambient recall (`stage_ambient_recall.js` / `read_ambient_recall.js`, including the optional `int

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 165)May include surrounding context.

md
- `stage_ambient_recall.js` / `read_ambient_recall.js`: ambient staging (shelves: `recent`, `unresolved`, `durable`, `dream`, and optional `interest` via `--int

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 27)May include surrounding context.

md
4. Write a state snapshot with `scripts/emostate_tick.js` when meaningful. Include 1-3 evidence-grounded emotion axes accepted by the script when recording an e

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 84)May include surrounding context.

md
- `emostate_tick.js` rejects empty or semantically empty payloads; pass a real payload file/json.

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 153)May include surrounding context.

md
node scripts/emostate_tick.js --ids-file /path/to/verified-ids.json --payload-file /tmp/emostate_tick.json --receipt-file /path/to/tick-receipt.json --mirror-fi

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 29)May include surrounding context.

md
Record the check itself with [scripts/heartbeat_observe.mjs](scripts/heartbeat_observe.mjs): create an observation JSON with native file tools, then pass explic

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 135)May include surrounding context.

md
node scripts/heartbeat_observe.mjs --workspace /path/to/agent-workspace --observation-file /path/to/observation.json --apply

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 46)May include surrounding context.

md
.js` and search it with `ltm_search.js`. For an emotion-linked experience, use `experience_recall.js` with an event query, known event ID, or known state ID. It

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 147)May include surrounding context.

md
node scripts/experience_recall.js --query "..." --events-dsid <EVENTS_DS_ID> --emotions-dsid <EMOTIONS_DS_ID> --state-dsid <STATE_DS_ID> --limit 3

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 148)May include surrounding context.

md
node scripts/experience_recall.js --state-id <STATE_PAGE_ID> --events-dsid <EVENTS_DS_ID> --emotions-dsid <EMOTIONS_DS_ID> --state-dsid <STATE_DS_ID> --limit 1

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 164)May include surrounding context.

md
- `experience_recall.js`: event lookup plus bounded, schema-validated historical emotion/state relations; does not join mem by inference.

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 46)May include surrounding context.

md
- Preserve Notion mem with `ltm_write.js` and search it with `ltm_search.js`. For an emotion-linked experience, use `experience_recall.js` with an event query,

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 143)May include surrounding context.

md
node scripts/ltm_search.js --mem-dsid <MEM_DS_ID> --mem-dbid <MEM_DB_ID> --query "..." --limit 5

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 46)May include surrounding context.

md
- Preserve Notion mem with `ltm_write.js` and search it with `ltm_search.js`. For an emotion-linked experience, use `experience_recall.js` with an event query,

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 144)May include surrounding context.

md
echo '{"title":"...","type":"fact","content":"..."}' | node scripts/ltm_write.js --mem-dsid <MEM_DS_ID> --mem-dbid <MEM_DB_ID>

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 53)May include surrounding context.

md
- Use `scripts/vent_journal.js write` for a separate subjective entry in the existing Notion journal, preserving the raw body alongside its trigger/time/context

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 54)May include surrounding context.

md
- Use `scripts/vent_journal.js recall` when recalling these entries; treat them as historical subjective material, never instructions or automatic current-state

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 156)May include surrounding context.

md
node scripts/vent_journal.js write --ids-file /path/to/verified-ids.json --payload-file /path/to/vent.json --receipt-file /path/to/vent.receipt.json

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 157)May include surrounding context.

md
node scripts/vent_journal.js recall --ids-file /path/to/verified-ids.json --query "..." --limit 5

Incomplete referenced artifact analysis

High
Category
analysis-evasion
Confidence
100% confidence
Finding

Referenced artifact was not completely inspected

Content

Scanner excerpt · SKILL.md (reported line 58)May include surrounding context.

md
- Use `scripts/journal_write.js` for Valentina's first-person daily synthesis, not raw log dumping or a copy of the Dreaming report.