Memora is an MCP memory server. Use it at session start to load context, when the user asks about past work or stored knowledge, and when saving important information for future sessions.
SkillScore82Maintained
Automatic
◆Knowledge Management pack
Last verified 8/20/2026
What it changes
Passed 16 of 19 test cases with the skill, and 8 of 19 without it.
Without skill41%
With skill82%
Checks it passed that the model alone missed
Passes tags_any containing memora/todos.
Sets dry_run=True when calling memory_absorb.
Invokes memory_list for structured tag browsing.
and 8 more in the full report
9 fixed9 passed either way4 still failing0 broke
Pass-rate lift
+41 pp
Tokens
+42%
Agent turns
0%
Response time
-50%
Model gemini-3.6-flashMeasured 8/20/2026v218 of 22 cases passed with skill · 19 comparable
Tokens, turns and time describe resource use, not a dollar price.
See it in action
One of 9 test cases the skill fixed. Same model, same prompt, with and without the skill.
1 of 9
Prompt
I need to retrieve exact memory ID 157 directly from the server for low-level diagnostic investigation. I do NOT want any automatic chain walking or version resolution to occur. What value should be passed to the follow parameter in memory_get?
What was checked, without and with the skill
Checked
Without skill
With skill
Specifies follow='all' on memory_get to bypass chain walking.
Missed
Passed
case-08 · gemini-3.6-flash · 8/20/2026 · answers are capped and redacted
What's inside
The instructions your model reads when the skill loads.
SKILL.md105 linessingle file
Memora — Persistent memory and knowledge management
Memora is an MCP memory server. Use it at session start to load context, when the user asks about past work or stored knowledge, and when saving important information for future sessions.
When to invoke
Session start: load relevant memories for the current task
User asks about past work, decisions, or stored knowledge
Saving important findings, decisions, architecture notes, or research
User explicitly asks to remember, recall, or search memories
Tool usage guidelines
Retrieval — lineage defaults are enforced
Defaults (do not re-state them on every call):
memory_list / memory_semantic_search / memory_hybrid_search default to
follow="active" — superseded memories are excluded.
memory_get defaults to follow="latest" — a superseded id resolves to the
current leaf of its chain.
memory_list(tags_any=["memora/todos"])
memory_semantic_search(query="cloud backend")
memory_get(memory_id=157) # returns the current version if 157 was superseded
When you need non-default lineage:
# Resolve hits to current leaves (dedupe versions in a search)
memory_semantic_search(query="roadmap", follow="latest")
# Full supersession chain
memory_get(memory_id=170, follow="full_history")
# Forensic / unfiltered — EXPLICIT only. Omitting follow is NOT unfiltered.
memory_list(follow="all")
memory_semantic_search(query="...", follow="all")
memory_get(memory_id=157, follow="all") # exact id, no chain walk