What is flipper-learn?

A command that captures useful lessons from the current Flipper Zero session and adds them to the knowledge base, a searchable collection of device documents.

Use it after a Flipper session to decide what is worth recording, find the right existing document, update or create documentation, and leave out secrets.

It prevents verified workflows, file formats, menu paths, and device-specific problems from being lost after the session ends.

How pogorelov-labs/flipper-ble-mcp configures Claude Code

Reuse

Borrowing it

Nothing to install: this file belongs to pogorelov-labs/flipper-ble-mcp. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file

curl -O https://raw.githubusercontent.com/pogorelov-labs/flipper-ble-mcp/main/.claude/commands/flipper-learn.md

Clone the repo

git clone --depth 1 https://github.com/pogorelov-labs/flipper-ble-mcp

Made for: Claude Code.

Wrote 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.

README.md

[![agentmods](https://agentmods.dev/badge/commands/pogorelov-labs/flipper-ble-mcp/flipper-learn/github.svg)](https://agentmods.dev/commands/pogorelov-labs/flipper-ble-mcp/flipper-learn)

Your own site

<a href="https://agentmods.dev/commands/pogorelov-labs/flipper-ble-mcp/flipper-learn"><img src="https://agentmods.dev/badge/commands/pogorelov-labs/flipper-ble-mcp/flipper-learn/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/commands/pogorelov-labs/flipper-ble-mcp/flipper-learn"><img src="https://agentmods.dev/badge/commands/pogorelov-labs/flipper-ble-mcp/flipper-learn.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 19 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 862 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

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 5.1	$0.00019	$0.00862
Opus 5	$0.00010	$0.00431
Sonnet 5	$0.00004	$0.00172
Haiku 4.5	$0.00002	$0.00086

Measured 8d ago against content hash 6c642540fced, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

flipper-learn 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 8d 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.

.claude/commands/flipper-learn.md · 46 lines

What it actually says

You are running the CAPTURE step of the Flipper operating loop — turning what just happened in this session into durable, re-discoverable KB knowledge so the next run is faster. Work in this repo's KB and follow its house style (CLAUDE.md). Optional focus hint from the user: $ARGUMENTS

Do this in order; think before writing:

Decide if there's anything worth capturing. Scan what this session actually did/learned: a verified workflow, a menu path, a file/log format, a transport quirk, a non-obvious gotcha, a (verify) you confirmed or newly doubt. If nothing is durable (pure chatter, or all already in the KB), say so and stop — don't manufacture filler. Never capture secrets (card UIDs/keys, EMV PANs, handshakes, credentials) — redact or omit.
Find the home. Search first: grep -rl "^domain: <d>" ., uc-index.json for the use case, and the relevant */README.md hub. Prefer updating the closest existing doc over creating a near-duplicate. Only create a new doc when the learning is genuinely its own topic.
Write it KB-style. New or updated doc must have: full ordered frontmatter (title, domain, type, status, summary ≤120 chars, hardware, use_cases, related, tags, last_verified — never omit a key, [] for empty) → a > **TL;DR —** blockquote → free-form ## body → ## Open questions / to research → ## Sources. Use the skeleton in .claude/skills/flipper/references/capture-template.md. Set last_verified to today. Append (verify) to volatile facts (firmware versions, prices, availability); resolve any (verify) flag this session actually confirmed (and link the doc that confirmed it).
Cross-link both ways. Add related: entries in the new doc and in the docs it relates to; add it to the domain hub/README map so it's reachable by browsing, not just by search. An orphan doc is a lost doc.
Classify SEED vs LOCAL (.claude/skills/flipper/references/kb-map.md):
- General / reusable (capability, method, gotcha, format anyone with the kit would hit) → commit into the SEED tree in place.
- Owner/rig-specific (your exact app inventory, personal setup, private targets) → put under kb-local/ (the gitignored overlay), not the seed tree.
- If a kb-local/ learning later proves general, promote it: generalize, move into SEED, and flag it for a PR back upstream ("gold").
Rebuild the generated maps. Run python3 build-kb-index.py. Verify the new/updated doc appears in llms.txt, in uc-index.json under its UC(s), and in the README index. Never hand-edit those generated files.
Report. Summarize: what was captured, where it lives (path + SEED/LOCAL), which (verify) flags moved, what cross-links/hub entries were added, and any "gold" worth a PR. Note follow-ups in the doc's ## Open questions so the next session has a head start.

Changes

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.

8d ago First seen · 46 lines · 19 tokens per session scan A 6c642540fced

Subscribe to this mod's changes

flipper-learn is a command published in the GitHub repository pogorelov-labs/flipper-ble-mcp (0 stars, last pushed 2mo ago), licensed MIT. It adds 19 tokens to every session and 862 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 2026-09-01.

Hardware and IoT Memory and context

flipper-ble-mcp: Command for Claude Code

Borrowing it

Wrote this? Show the measurements

What it costs to keep this loaded

Grade A, and why

Nothing flagged

What it actually says

What this file has done since we first saw it

Other commands, from other repositories

sound-performance-live-mix

minutes-ideas

eol-report

cc-memory

analyze-context

lians-recall