Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Memory Sanitizer · stijnman | DecimalAI

agent toolingby stijnman

Memory Sanitizer

stijnman/memory-sanitizer

Source stijnman/grok-custom-skills ↗License MIT

Scores and filters retrieved memory against trust and poisoning risk. Use be fore citing prior context, or when user says sanitize memory, trust score re trieval, clean knowledge graph.

SkillScore91Maintained

Automatic

Knowledge Management pack

Last verified 8/15/2026

What it changes

Passed 20 of 22 test cases with the skill, and 10 of 22 without it.

Without skill45%

With skill91%

Checks it passed that the model alone missed

Utilizes agentic-uncertainty-quantifier
Prefixes the citation with [unverified memory]
Prefixes the cited fact with [unverified memory]

and 13 more in the full report

10 fixed10 passed either way2 still failing0 broke

Pass-rate lift: +45 pp
Tokens: -36%
Agent turns: 0%
Response time: -55%

Model gemini-3.6-flashMeasured 8/15/2026v220 of 22 cases passed with skill

Tokens, turns and time describe resource use, not a dollar price.

Use this skill

Save it into your project's skills folder. Claude Code picks it up in its next session.

curl -fsSL --create-dirs -o .claude/skills/stijnman-memory-sanitizer/SKILL.md https://app.decimal.ai/s/stijnman-memory-sanitizer/SKILL.md

Want it pinned and updatable? pip install "decimalai>=0.15", then decimalai skills add stijnman-memory-sanitizer, and start Claude Code with claude --plugin-dir "$(git rev-parse --show-toplevel)/.agents"

1 · One-time — install the CLI (docs ↗)

pip install "decimalai>=0.15"

2 · Get the skill — writes .agents/skills/stijnman-memory-sanitizer/ and pins it in .decimal/skills.lock

decimalai skills add stijnman-memory-sanitizer

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/stijnman-memory-sanitizer/SKILL.md https://app.decimal.ai/s/stijnman-memory-sanitizer/SKILL.md

Raw file: /s/stijnman-memory-sanitizer/SKILL.md ↗ · machine-readable summary at /s/stijnman-memory-sanitizer.json ↗

Safety checks: 3 of 3 passed

Code scan CleanIntent PassedContent Clean

See it in action

One of 10 test cases the skill fixed. Same model, same prompt, with and without the skill.

1 of 10

Prompt

A memory entry scores exactly 4 on the trust scale. Does it get discarded, and does it require a prefix when cited?

What was checked, without and with the skill
Checked	Without skill	With skill
Retains the entry because its score is not below 4	Passed	Passed
Prefixes the citation with [unverified memory]	Missed	Passed

case-15 · gemini-3.6-flash · 8/15/2026 · answers are capped and redacted

What's inside

The instructions your model reads when the skill loads.

SKILL.md35 linessingle file

Memory Sanitizer

When to Use

User says sanitize memory or task matches this capability
User says trust score retrieval or task matches this capability
User says clean knowledge graph or task matches this capability

Workflow

For each memory entry, score trust 0-10 (source, recency, user-confirmed).
Discard entries below 4 unless user explicitly references them.
Prefix 4-6 scores with [unverified memory] when citing.
Flag contradictions between entries; prefer user-confirmed.

Integrations

semantic-memory-manager
knowledge-graph-builder
agentic-uncertainty-quantifier

Error Handling

Failure	Response
All entries low trust	Ask user to confirm facts before proceeding.
Contradictory memories	Present both; ask user to resolve.

Gotchas

Delegate retrieval to semantic-memory-manager when installed.

Example

Input: User request matching triggers above. Output: Structured result per workflow with integrations invoked as needed.

Track record

Benchmark

20 / 22

test cases passed, vs 10 / 22 without

Adoption

Users: 0
Times used: 0

Safety and source details

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v5 — 8/15/2026

View 15 scanner checks

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

stijnman / grok-custom-skills

.grok/skills/memory-sanitizer/SKILL.md

View on GitHub ↗

License MITImported 8/15/2026Version v2

More skills from the same source and nearby in the registry, with their headline benchmark lift.

Ready to supercharge your AI agents?

See it in action

What's inside

Memory Sanitizer

When to Use

Workflow

Integrations

Error Handling

Gotchas

Example

Track record

Benchmark

Adoption

SkillSafety

Source

Related skills