memory-exchange
Skill617 starsChanged 10 months agoMITView on GitHub
Safely export, inspect, stage, analyze, review, and decide protected replacements for Hexis Memory Exchange files
We split this skill into the rules it gives an agent
14 rules. About 0 are common, 5 are distinctive, and 9 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈5
Distinctive · estimate
This author's own choices. Worth a look.
9
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 14
Use this skill to move memory between Hexis instances or inspect another instance's memory without silently blending it into active state.
- Very specific: line 20
Use export_memories with the narrowest useful section, time, and redaction scope.
- Very specific: line 29
Run import_dry_run first. Report whether import is permitted, conflicts, protected-state policy, warnings, and estimated embedding work.
- Very specific: line 34
Call import_memories only after the user confirms the file's exact declared intent.
- Distinctive: line 41
Do not describe imported memories as semantically searchable until the import result or later re-embedding workflow confirms that they are …
- Very specific: line 49
Use import_review to inspect pending records and conflicts.
- Very specific: line 52
Use import_reject with a reason, import_modify with a material-change kind and reason, or import_quote to retain foreign context as …
- Very specific: line 62
Use protected_replacement_list to retrieve every open request, then use protected_replacement_inspect to compare the actual imported …
- Very specific: line 65
Use protected_replacement_review to accept, refuse, request modifications, or defer.
- Very specific: line 73
Use protected_replacement_audit_list with a narrow time range to inspect immutable local replacement, verification, and reversion history.
- Very specific: line 82
Use protected_reversion_list to find executed replacements whose earlier-of heartbeat and wall-clock windows remain open.
- Distinctive: line 94
Treat HMX files as sensitive user data.
- Distinctive: line 95
Never infer consent for protected sections, raw material, or file overwrite.
- Distinctive: line 96
Keep analysis-only records outside active recall until explicitly promoted, reviewed, and accepted.
≈ “Common” is an estimate: at least 1 in 5 of the 308,232 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Hexis Memory Exchange
Use this skill to move memory between Hexis instances or inspect another instance's memory without silently blending it into active state.
Export
- Establish the user's intent:
port,duplicate,telepathy, oranalysis. - Use
export_memorieswith the narrowest useful section, time, and redaction scope. Protected state, raw units, configuration, in-flight work, and audit records require an explicit request. - Omit
output_pathwhen returning the exchange in place. For file output, use a workspace path and never setoverwriteunless the user explicitly chose to replace that file.
Import
- Run
import_dry_runfirst. Report whether import is permitted, conflicts, protected-state policy, warnings, and estimated embedding work. - Derive the default strategy from the file's intent unless the user chose a different supported strategy. Telepathy defaults to deliberative staging; analysis stays isolated.
- Call
import_memoriesonly after the user confirms the file's exact declared intent. Skipping identity, worldview, or narrative is available when the user wants a narrower import. - For
portorduplicateinto an active target, authoritative import requires explicitreplace_sectionsandreplacement_rationale. Unselected protected sections remain unchanged. A successful request may still await the agent's acknowledgement; report its replacement IDs instead of claiming completion. - Do not describe imported memories as semantically searchable until the import result or later re-embedding workflow confirms that they are ready.
- Preserve failed in-flight work as diagnostics by default. Set
retry_failed_workonly when the user explicitly chooses to rerun those failed consolidation or reconsolidation tasks.
Deliberative Review
- Use
import_reviewto inspect pending records and conflicts. - Decide one record at a time. Use
import_acceptonly when the record should become active; protected active-state policy cannot be bypassed. - Use
import_rejectwith a reason,import_modifywith a material-change kind and reason, orimport_quoteto retain foreign context as archived evidence. promote_to_stagedcopies an analysis record into review without copying its embedding.demote_to_analysismoves a pending staged record back into isolation. Both require a rationale.
Protected Replacement Review
- A pending protected replacement is a request, not permission to mutate identity, worldview, goals, drives, emotional triggers, or narrative.
- Use
protected_replacement_listto retrieve every open request, then useprotected_replacement_inspectto compare the actual imported section with current local state and check whether local state changed after the request. Useprotected_replacement_reviewto accept, refuse, request modifications, or defer. Acceptance executes the snapshot, immutable audit, whole-section replacement, and digest verification atomically; a failure leaves the request pending. - Refusal and modification requests require a rationale. Modification requests
also require concrete
proposed_changes. Do not accept merely because source and target claim the same lineage; content-identical verified operations never enter this queue. - Use
protected_replacement_audit_listwith a narrow time range to inspect immutable local replacement, verification, and reversion history. Imported foreign diagnostics are not reported as local history. - Operator override is not an agent tool and cannot be requested through this skill. A refusal or modification request is a completed decision, not an invitation to bypass acknowledgement.
Protected Replacement Reversion
- Use
protected_reversion_listto find executed replacements whose earlier-of heartbeat and wall-clock windows remain open. Reversion never runs on a timer. - Inspect the replacement before reverting. Use
protected_replacement_revertwith the replacement audit ID and a concrete rationale only when restoring the snapshot is the intended choice. - Reversion refuses to overwrite protected state that changed after the replacement. It atomically restores and verifies the snapshot, writes an immutable reversion audit, then purges the consumed payload while retaining its tombstone. A failed restore leaves the window and current state intact.
Boundaries
- Treat HMX files as sensitive user data.
- Never infer consent for protected sections, raw material, or file overwrite.
- Keep analysis-only records outside active recall until explicitly promoted, reviewed, and accepted.
- On failure, preserve the file and staged records and give the exact corrective action; do not retry with broader scope automatically.
Other versions1 project uses a version of this.
Word for word · 1
- QuixiAI/Hexis★ 617100% same
Edited · 0
None found.
Wrote their own version · 0
None found.
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository42 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- audio-analysisSkill
- calendarSkill
- code-executionSkill
- connector-action-authorizationSkill
- core-memorySkill
- cost-reportSkill
- councilSkill
- crm-lookupSkill
- daily-briefingSkill
- delegationSkill
- email-digestSkill
- expense-captureSkill
- github-issuesSkill
- gmail-actionsSkill
- gmail-connector-setupSkill
- home-assistantSkill
- host-nodeSkill
- humanizerSkill
- image-genSkill
- inbox-triageSkill
- integration-connector-setupSkill
- knowledge-ingestSkill
- meeting-follow-upSkill
- meeting-prepSkill
- notionSkill
- operationsSkill
- outreachSkill
- personal-notesSkill
- researchSkill
- self-inspectionSkill
- self-reflectionSkill
- skill-authoringSkill
- speechSkill
- spotifySkill
- travel-prepSkill
- trelloSkill
- twitter-x-actionsSkill
- twitter-x-connector-setupSkill
- weatherSkill
- weekly-reviewSkill
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
Reviews can't be read right now.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

