Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add commands/goldenwo/universal-memory/um-previewgit clone --depth 1 https://github.com/goldenwo/universal-memoryWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/goldenwo/universal-memory/um-preview)<a href="https://agentmods.dev/commands/goldenwo/universal-memory/um-preview"><img src="https://agentmods.dev/badge/commands/goldenwo/universal-memory/um-preview.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00014 | $0.00206 |
| Opus 5 | $0.00007 | $0.00103 |
| Sonnet 5 | $0.00003 | $0.00041 |
| Haiku 4.5 | $0.00001 | $0.00021 |
Grade A, and why
um-preview scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
Execute bash "${CLAUDE_PLUGIN_ROOT}/bin/um-preview" via the Bash tool and show the output verbatim.
This is a PREVIEW — a draft of what state.md would contain if session-end.sh ran right now. It does NOT modify any files in the vault:
- No write to
$UM_VAULT_DIR/state/<project>/state.md - No lockdir acquisition
- No cost-log.csv telemetry append (
update-state.sh --stdoutsuppresses it) - No reindex
Use this to sanity-check the merge before committing. If the preview looks good and you want to commit it, run /um-checkpoint instead — that invokes the canonical session-end pipeline that writes to disk and reindexes.
The underlying CLI is fail-soft: missing API key, empty captures, or LLM errors all exit non-zero with a human-readable message on stderr; they never corrupt data.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 4d ago First seen · 16 lines · 14 tokens per session scan A e1f2d7ac73b1
um-preview is a command published in the GitHub repository goldenwo/universal-memory (2 stars, last pushed 6d ago), licensed MIT. It adds 14 tokens to every session and 206 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
capture-answer
Ingests an expert's answers to an L2 questionnaire in the abapwiki knowledge base: transforms the answers (filled in the interviews/ file or given verbally) into a canonical citable expert-answer (slices/ /inputs/expert-answers/), moves gaps to 'answered' and closes the questions. Use this skill when the owner/expert…
ingest-l1
L1 analysis loop for the abapwiki knowledge base: for each batch it launches the abap-analyzer sub-agent in parallel, then the adversarial judge abap-deepcheck (separate session), applies only the analyses that pass the fail-closed gate, and commits. Resumes exactly after an interruption. Use this skill to document…
research-l2
Executes Phases 1-3 of the L2 process of the abapwiki knowledge base on a slice: launches the abap-functional-researcher sub-agent (gap discovery + multi-source auto-research: wiki -> raw/docs -> MCP abap-fs on the system read-only -> standard knowledge), ingests gaps and evidence, and generates questionnaires for…
slice-init
Creates an L2 slice in the abapwiki knowledge base: a business process on which to build functional documentation. Writes the manifest (curated anchors + mandatory real owner) and derives the membership from the dependency graph (BFS hop<=2). Use this skill to start the L2 process on a custom domain. Does not promote…
ingest-l0
L0 bootstrap of the abapwiki knowledge base: creates stubs for all custom objects from the TADIR. Use this skill to initialize the repo from scratch or to ingest a new TADIR export. Deterministic operations (no LLM sub-agent), idempotent.
lint-audit
Checks the integrity of the abapwiki knowledge base: parseable YAML frontmatter, no broken wikilinks, resolvable [VERIFIED: path:N-M] citations, non-nested confidence tags, drift between wiki and DB, synchronization between canonical agent contracts and the .claude/agents/ copies. Use this skill periodically or after…