BB omem
Search the user's local work memory (Office Memory / OMem) — a unified, locally-indexed wiki built from the user's emails, calendar events, documents, and collaboration notes. OMem ingests from whatever sources the user has configured: mail kind (e.g. macOS Mail.app, Outlook Classic, Outlook on the Web, AppleScript-driven mailboxes), calendar kind (Calendar.app, Outlook calendars, .ics files), file kind (any local folder — OneDrive, Dropbox, iCloud Drive, ~/Documents, plain directories), and loop kind (Microsoft Teams Loop notes). Future versions add Slack / Jira / Notion. The wiki abstracts over all of these — you don't need to know which source a hit came from. USE THIS SKILL whenever the user's question touches their work context, even implicitly — past meetings, colleagues' names, project codenames, vendor/contract details, "that document I saw last week", "what did X say about Y", upcoming/past calendar items, internal acronyms, RFPs, performance reviews, OKRs, anything that sounds like it could live in their work history (any inbox, calendar, local folder, or collaboration tool they use for work). When in doubt, CALL omem query FIRST and decide afterward whether the results help. DO NOT use this skill for: weather, general programming questions, translation, math, public-knowledge lookups, or anything where the user clearly is NOT asking about their own work history. Primary tool: `omem query "<question>" --format json --limit 20` Then progressively drill down via `omem page get <page_id>`, `omem raw get <page_id> --parsed`, `omem raw get <page_id>`. Usually one query and one page answers the question — read it, cite, done. Only when a page's answer hinges on a specific term, person, or document you haven't resolved, query again for that one thing — follow the thread, don't crawl the graph. Never call `omem setup / install / ingest / lint / index rebuild` — those are user actions; tell the user in natural language instead.
Search the user's local work memory (Office Memory / OMem) — a unified, locally-indexed wiki built from the user's emails, calendar events, documents, and…
As a process B 78/100 · Nearly there — weak spots: result and completion, running it twice
What is at stake
Medium-severity findings: the skill is probably honest, but read what alarmed the scanner.
Below is the worst case for this category. The finding here is medium: the guard saw a sign, not a proof.
The skill asks for more than the task needs: broad tool access, credential environment variables, binaries. Every extra permission widens the damage from a mistake or a compromise.
Narrow allowed-tools and the variable list to the minimum; replace binaries with readable sources or scripts.
How to improve
- 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 · 2
✓ No critical or high findings
Medium and low: 2
-
medium Broad scope
meta-broad-allowed-toolsSKILL.md:1Broad tool permissions pre-approved: Bashallowed-tools: Bash Read
-
low Dangerous commands
cmd-pipe-to-shell-known-hostreferences/troubleshooting.md:14Pipe-to-shell installer from a well-known host (still executes remote code) (quoted — discussed, not commanded)> `curl -fsSL https://github.com/seacen/omem/releases/latest/download/install.sh | sh`
quoted
Files scanned: 5. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- error
description-longdescription is 1966 chars, limit 1024 - note
description-budgetdescription takes 1966 of the ~15000-char shared budget for all skills
Process rating: all ten parameters 78/100
- 0Result and completion. Does not say what the result is
- 30Running it twice. 4 mutating operations with no state check
- 70Inputs and preconditions. Inputs and preconditions are listed
- 85Steps. 65 steps, 3 vague phrases
- 100Tools and files. Tools declared in frontmatter
- 100When it triggers. States when to use and when not to
- 100Failures and branches. 3 branches, has a failure section
- 100Consistency. Name and required fields are in place
- 100Execution cost. Instruction body is 3818 tokens
- 100Progress reporting. Reports progress
- low 10 top-level sections: this looks like several domains in one skill
- low The skill ranks results itself: that belongs to the system behind the tool, not the model
- low The response is described with custom markup (4 tags): a typed call is more reliable
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 1965: 120–800 characters recommended
- +3Output format is not stated: the model decides each time
- +1No license
- +2Single-language instructions
- +5Description quotes 3 example trigger phrases
- +4Description says when NOT to use the skill
- +4Structure: 16 headings
- +3Step-by-step instructions: 65 items
- +4Has examples (1 code blocks)
- +4Reference files are cited in the instructions (3 of 3)
Quality base 70; lint remarks subtract, signals add up to 100. Result: 66.