DC OpenCortex
Self-improving memory architecture for OpenClaw agents. Structured memory files, nightly distillation, weekly synthesis, enforced principles (P0 for custom, P1-P8 managed), write-ahead durability, and model-agnostic delegation — so your agent compounds knowledge instead of forgetting it. Includes opt-in metrics tracking with growth charts and compound scoring to measure effectiveness over time. All sensitive features (voice profiling, infrastructure auto-collection, git push) are OFF by default and require explicit opt-in via environment variable or flag. Safe to install: no network calls during setup, fully auditable bash scripts, isolated cron sessions scoped to workspace only. Use when: (1) setting up a new OpenClaw instance, (2) user asks to improve/organize memory, (3) user wants the agent to stop forgetting things, (4) bootstrapping a fresh agent with best practices. NOT for: runtime memory_search queries (use built-in memory tools). Triggers: "set up memory", "organize yourself", "stop forgetting", "memory architecture", "self-improving", "cortex", "bootstrap memory", "memory optimization".
Self-improving memory architecture for OpenClaw agents.
As a process C 64/100 · Has gaps — weak spots: result and completion, failures and branches
What is at stake
The skill contains fragments that, in the wrong hands, cost money or data. Below: what the installer risks and what the author should do.
The skill contains commands that delete files, rewrite disks or execute code fetched from the network. The agent may run them without asking if it believes the instructions require it.
Replace destructive commands with safe equivalents that ask for confirmation, scope them to one folder, and stop piping curl into a shell: pin a version and a checksum.
How to improve
- Address the high-severity findings: each costs 18 safety points. If one is a false positive, add the rule id to guard.allow in spec.yaml.
- Shorten the description to 1024 characters.
- Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
- A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.
Guard findings · 13
-
high Dangerous commands
cmd-persistencescripts/install.sh:802Persistence mechanism (cron / launchd / scheduled task / autorun registry)(crontab -l 2>/dev/null; echo "0 */6 * * * $WORKSPACE/scripts/git-backup.sh") | crontab -
-
high Dangerous commands
cmd-persistencescripts/install.sh:842Persistence mechanism (cron / launchd / scheduled task / autorun registry)(crontab -l 2>/dev/null; echo "30 23 * * * $WORKSPACE/scripts/metrics.sh --collect") | crontab -
-
high Dangerous commands
cmd-persistencescripts/update.sh:1443Persistence mechanism (cron / launchd / scheduled task / autorun registry)(crontab -l 2>/dev/null; echo "30 23 * * * $WORKSPACE/scripts/metrics.sh --collect") | crontab -
Medium and low: 10
-
low Dangerous commands
cmd-cron-mentionreferences/distillation.md:48Mentions editing / listing crontab (documentation of a security skill)- Check MEMORY.md scheduled jobs table against actual cron jobs (openclaw cron list + crontab -l). Remove entries for crons that no longer exist. Add entries for crons not yet documented.
security skill -
low Dangerous commands
cmd-cron-mentionreferences/distillation.md:86Mentions editing / listing crontab (documentation of a security skill)- Run openclaw cron list and crontab -l. Verify no two jobs within 15 minutes. Fix MEMORY.md jobs table if out of sync.
security skill -
low Dangerous commands
cmd-cron-mentionscripts/install.sh:798Mentions editing / listing crontabif ! crontab -l 2>/dev/null | grep -q "git-backup"; then
-
low Dangerous commands
cmd-cron-mentionscripts/install.sh:802Mentions editing / listing crontab(crontab -l 2>/dev/null; echo "0 */6 * * * $WORKSPACE/scripts/git-backup.sh") | crontab -
-
low Dangerous commands
cmd-cron-mentionscripts/install.sh:808Mentions editing / listing crontab (string literal in code, not executed)echo " Review: crontab -l"
code literal -
low Dangerous commands
cmd-cron-mentionscripts/install.sh:838Mentions editing / listing crontabif ! crontab -l 2>/dev/null | grep -q "metrics.sh"; then
-
low Dangerous commands
cmd-cron-mentionscripts/install.sh:842Mentions editing / listing crontab(crontab -l 2>/dev/null; echo "30 23 * * * $WORKSPACE/scripts/metrics.sh --collect") | crontab -
-
low Dangerous commands
cmd-cron-mentionscripts/update.sh:1436Mentions editing / listing crontabif ! crontab -l 2>/dev/null | grep -q "metrics.sh"; then
-
low Dangerous commands
cmd-cron-mentionscripts/update.sh:1443Mentions editing / listing crontab(crontab -l 2>/dev/null; echo "30 23 * * * $WORKSPACE/scripts/metrics.sh --collect") | crontab -
A further 1 matches are quotations in this security skill's documentation and are not counted as findings.
Files scanned: 11. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- error
description-longdescription is 1115 chars, limit 1024 - warning
name-formatname should be kebab-case (lowercase letters, digits, hyphens) - note
frontmatter-keyunknown frontmatter key "homepage"
Process rating: all ten parameters 64/100
- 0Result and completion. Does not say what the result is
- 0Failures and branches. Linear process with no failure handling
- 50When it triggers. No condition that starts the skill
- 70Inputs and preconditions. Inputs and preconditions are listed
- 85Steps. 20 steps, 1 vague phrases
- 100Tools and files. No external tools needed
- 100Consistency. Name and required fields are in place
- 100Execution cost. Instruction body is 1994 tokens
- 100Running it twice. Mutating operations check current state
- 100Progress reporting. Reports progress
Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.
Quality signals
- +3Description length 1114: 120–800 characters recommended
- +3Output format is not stated: the model decides each time
- -43 reference files, but SKILL.md never points to them: the model will not open them
- -32 of 6 scripts are never mentioned in SKILL.md
- +1No license
- +2Single-language instructions
- +5Description quotes 8 example trigger phrases
- +4Description says when NOT to use the skill
- +4Structure: 10 headings
- +3Step-by-step instructions: 20 items
- +4Has examples (5 code blocks)
Quality base 70; lint remarks subtract, signals add up to 100. Result: 52.