SageMaker

Unified memory-and-growth operating system for agents. Use when you need consistent layered memory (short/mid/long/knowledge), self-model-driven promotion ru...

Security Scan

The skill's instructions, file usage, and lack of requested credentials or installs are consistent with a local 'agent memory' helper; no clear mismatches were found, but some operational ambiguities merit caution.

Purpose & Capability

The name/description describe a layered memory system and the SKILL.md only asks the agent to create/read/update local memory files and governance documents (memory/, self-model.md, etc.), which is coherent with the stated purpose.

Instruction Scope

Instructions stay within the memory/ and governance-file scope and prescribe promotion rules, gating, and pre/post task behavior. They do reference optionally running skills/SageMaker/scripts/install.ps1 if present; the skill bundle currently has no scripts, so there is no immediate executable action. The SKILL.md does not explicitly constrain the root workspace for the memory/ paths, so where files will be created/read is ambiguous and worth confirming.

Install Mechanism

No install spec or bundled code is present (instruction-only), so nothing will be written or downloaded by default. The doc mentions an optional PowerShell installer if bundled; because no install script is included, there is no install-time risk in this package as provided.

Credentials

The skill requests no environment variables, credentials, or config paths. Its file-read/write behavior is proportionate to a memory-management helper.

Persistence & Privilege

always:false and standard autonomous invocation are used. The skill expects persistent files (memory/...), which is appropriate for a memory OS, but you should confirm the intended storage location and retention policy before use. There is no request to modify other skills or system-wide settings.

Assessment

This skill is internally consistent with a local agent memory管理 helper and doesn't request credentials or installs. Before enabling it: 1) Confirm where the agent will read/write the memory/ and governance files (workspace root vs system paths) so sensitive data isn't stored unexpectedly. 2) If an installer (scripts/install.ps1) is ever bundled, inspect the script before running it — do not run PowerShell scripts from untrusted sources or with execution-policy bypass. 3) Decide retention and privacy rules for stored memories (they can contain sensitive info). 4) If you expect the skill to integrate across multiple agents or systems, request details on access controls. If those questions are answered satisfactorily, the skill appears safe to use.

Neuro Memory Core

Implement one shared behavior loop:

experience -> short_term evidence -> mid_term synthesis -> (knowledge or long_term) -> self-model calibration -> next-task behavior

Install / Bootstrap

Preferred safe path (no execution-policy bypass): create/verify these artifacts directly:

• memory/short_term/
• memory/mid_term/MEMORY.md
• memory/long_term/MEMORY.md
• memory/knowledge.md
• memory/check_memory.json (dual-gate schema)

If this skill bundle includes scripts/install.ps1, you may run it without bypassing execution policy:

powershell -File "skills/SageMaker/scripts/install.ps1"

Optional (if bundled installer supports it): apply HEARTBEAT template with backup:

powershell -File "skills/SageMaker/scripts/install.ps1" -ApplyHeartbeatTemplate

Canonical Files

• memory/short_term/YYYY-MM-DD.md (raw evidence)
• memory/mid_term/MEMORY.md (near-term reusable conclusions)
• memory/long_term/MEMORY.md (stable collaboration constraints)
• memory/knowledge.md (transferable methods/policies)
• self-model.md (current strengths/failures/growth themes/uncertainties)
• memory/check_memory.json (daily/weekly gate state)

Required Promotion Rules

Update self-model.md only when:

1. recurring failure mode is re-validated
2. strength gains new evidence
3. active growth theme should switch

Update memory/knowledge.md only when:

1. method is reusable across multiple scenarios
2. explicit evidence supports it
3. rule can be expressed as if X then Y

Update memory/long_term/MEMORY.md only when:

1. improves long-term collaboration quality
2. is not short-term fluctuation

Entry Quality Contract (mid/long)

Every promoted item must include:

• reason
• evidence
• confidence (low|medium|high)

Task Coupling (Mandatory)

For medium/high complexity tasks:

1. read self-model.md + memory/knowledge.md
2. write preflight checklist:
   • goal
   • success criteria
   • risks
   • uncertainty
3. include one post-task reflection item before execution starts

Gate Model (memory/check_memory.json)

Use dual gate state:

{
  "daily_need_update": 1,
  "daily_update_done": 0,
  "weekly_need_update": 1,
  "weekly_update_done": 0
}

Semantics:

• 1/0 = pending
• 0/1 = done
• anything else = invalid; normalize to 1/0

Scheduling Pattern

• Daily cycle: promote short->mid + prune short-term retention
• Weekly cycle: promote mid->long
• Heartbeat: recovery path only (when gate remains pending)

Strict rule: success is valid only if gate flips to done.

Core-file Safety

Core-file changes must be proposal-first:

• draft proposal
• get approval
• then apply

Core files include: SOUL.md, IDENTITY.md, AGENTS.md and equivalent identity/behavior governance files.