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/backnotprop/plannotator/plannotator-lastgit clone --depth 1 https://github.com/backnotprop/plannotatorWhat 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 | $0.00007 | $0.00235 |
| Opus 5 | $0.00003 | $0.00118 |
| Sonnet 5 | $0.00001 | $0.00047 |
| Haiku 4.5 | $0.00001 | $0.00023 |
Grade A, and why
plannotator-last 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 yesterday.
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
Message Annotations
!plannotator copilot-last $ARGUMENTS
Your task
The output above will be one of:
- The exact text
The user approved., OR a JSON object with"decision": "approved". The user approved your last message. If that object also carries a"feedback"field, the user approved with notes: read them and carry them into subsequent work — they are non-blocking guidance, not a request to revise the message. Otherwise acknowledge with a single sentence ("Approved.") and stop. Either way, do not begin any work. - Empty, OR a JSON object with
"decision": "dismissed". The user closed the session without requesting changes. Acknowledge with a single sentence ("Annotation session closed.") and stop. Do not begin any work. - Plaintext annotation feedback, OR a JSON object with
"decision": "annotated"and a"feedback"field. Address the feedback. The user has reviewed your last message and provided specific annotations and comments.
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.
- yesterday First seen · 17 lines · 7 tokens per session scan A 237266b0fde8
plannotator-last is a command published in the GitHub repository backnotprop/plannotator (8,256 stars, last pushed 2d ago), licensed Apache-2.0. It adds 7 tokens to every session and 235 once invoked, about $0.0000 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 2026-08-30.
Other commands, from other repositories
kg-ingest
Ingest a source document into the Knowledge Graph - extract entities, concepts, create wiki pages.
kg-init
Initialize a new Knowledge Graph with raw/ and wiki/ structure.
kg-lint
Health check the Knowledge Graph - find orphans, contradictions, gaps, stale claims.
kg-project
Create a new project in the Knowledge Graph with ADR tracking.
kg-query
Query KG (supports: active, recent, search , decisions , plan ).
kg-setup
Configure Knowledge Graph path and create config file.