ffc-core
Skill16 starsChanged 37 days agoMITView on GitHub
Read and change Frappe or ERPNext data from the terminal with ffc (Foxmayn Frappe CLI) - get, list, count, create, update and delete documents, filters and fields, DocType schemas, output formats and --jq, exit codes, dry runs and safe-write habits. Use this whenever the user mentions ffc, a Frappe or ERPNext site, a DocType or a document (Sales Invoice, Customer, ToDo, Item...), or wants Frappe data in a script, even if they never name ffc. Start here; it routes to the other ffc-* skills.
In this repository13 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- desktop/CLAUDE.mdCLAUDE.md
- ffc-bulk-lifecycleSkill
- ffc-dev-clientSkill
- ffc-dev-commandSkill
- ffc-dev-desktopSkill
- ffc-dev-mcp-toolSkill
- ffc-devSkill
- ffc-dev-testingSkill
- ffc-files-collabSkill
- ffc-mcpSkill
- ffc-reports-apiSkill
- ffc-setupSkill
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 0 are distinctive, and 4 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 18
Use machine output: Add --json (-j) or --output json|ndjson|csv|tsv|yaml.
- Very specific: line 21
Never show secrets: Do not print config.yaml or any api_secret, password, token or the bearer token from ffc mcp status.
- Very specific: line 57
Avoid lost updates: read modified (get-doc --keys modified) and pass it to --if-unmodified.
- Very specific: line 60
Every flag, plus Single DocTypes, schema keys, whoami, can and ping: references/documents.md.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
ffc core: query and change Frappe data
ffc is a Go command-line client for Frappe and ERPNext sites over the REST API. It is a community project (Foxmayn), not an official Frappe project.
Before the first command
ffc --versionfails, orffc site listshows no site: use the ffc-setup skill.ffc site list --jsonlists the configured sites. With more than one, pass-s NAMEon every command so a write never lands on the default site by accident.ffc whoami --jsonnames the user, their roles and the Frappe major version (.server.major; some commands need v16).
Rules that keep an agent safe
- Use machine output. Add
--json(-j) or--output json|ndjson|csv|tsv|yaml. The table view is for humans: formatted numbers, cut columns. Machine output also turns errors into JSON on stderr. - Never wait on a prompt. Without a terminal ffc refuses to prompt (
--no-inputforces this) and fails with "pass --yes". Add--yesonly after the user approved that exact write. - Read, preview, then write.
get-schemafor field names and types,get-docfor current values,--dry-runto see the request (it sends no write and exits 0),ffc canto check the right. Ask the user before writing to any site that is not clearly a test site. - Never show secrets. Do not print
config.yamlor anyapi_secret,password, token or the bearer token fromffc mcp status.--debugredacts secrets; still keep--debug=bodyoutput out of shared logs. - Read the exit code. 0 ok, 2 usage, 3 auth, 4 not found, 5 permission, 6 validation or conflict, 7 network or server, 8 partial bulk failure. Details: references/exit-codes.md.
Global flags
| Flag | Meaning |
-s, --site NAME | Site from the config (default: default_site) |
-c, --config PATH | Config file (default ~/.config/ffc/config.yaml) |
-j, --json | JSON output (same as --output json) |
--output FMT | table, json, ndjson, csv, tsv, yaml |
--jq EXPR | Filter the JSON result with jq (strings print raw) |
--no-input | Never prompt; fail instead |
-q, --quiet | No spinner |
--timeout DUR | Per-request timeout, default 30s (raise for heavy reports) |
--debug[=body] | Trace HTTP on stderr, secrets redacted |
Env vars FFC_SITE, FFC_CONFIG, FFC_TIMEOUT, FFC_OUTPUT, FFC_DEBUG set the same things; a flag wins.
Documents
ffc get-doc -d "Sales Invoice" -n SINV-0001 --json --keys name,status,grand_total
ffc get-doc -d "System Settings" --json # Single DocType: --name optional
ffc list-docs -d ToDo -f name,status,modified --filters '{"status":"Open"}' -o "modified desc" -l 50 --json
ffc list-docs -d "Sales Invoice" --all --output csv -f name,customer,grand_total > invoices.csv
ffc count-docs -d "Sales Invoice" --filters '[["outstanding_amount",">",0]]' --json
ffc create-doc -d ToDo --data '{"description":"Call Acme"}' --json --keys name
ffc update-doc -d ToDo -n TD-0001 --data '{"status":"Closed"}' --dry-run --json
ffc update-doc -d ToDo -n TD-0001 --data '{"status":"Closed"}' --if-unmodified "2026-10-05 16:41:58.083711" --json
ffc delete-doc -d ToDo -n TD-0001 --dry-run --json # then --yes, once approvedlist-docsreturns onlynameunless you pass-f. Default limit 20;-l 0= no limit;--startoffsets;--allpages through everything.- JSON flags (
--filters,--data,--set,--args) also take@file.json, or@-for stdin. update-docsends only the fields in--data. A child table you send replaces the whole table, so send every row you want to keep.- Avoid lost updates: read
modified(get-doc --keys modified) and pass it to--if-unmodified; exit 6 means someone saved in between, so read again and redo. edit-docopens$EDITORand needs a human at a terminal. Agents useupdate-doc.
Every flag, plus Single DocTypes, schema keys, whoami, can and ping: references/documents.md.
Schema, names and permissions
ffc list-doctypes --module Accounts --json
ffc get-schema -d "Sales Invoice" --json --keys fields # compact field list
ffc get-schema -d "Sales Invoice" --json --refresh # skip the 1 h cache after Customize Form
ffc search acme -d Customer --json # title -> document name
ffc can -d "Sales Invoice" --perm create # exit 5 when denied
ffc can -d ToDo -n TD-0001 --perm write --jsonReference files
- references/filters-and-fields.md: filter syntax, operators, child-table filters, fields, ordering, paging.
- references/output-formats.md: formats,
--jq,--keys,--all, numbers, error JSON. - references/exit-codes.md: every exit code and when it happens.
- references/documents.md: flags of the document, schema and identity commands.
Other ffc skills
| Task | Skill |
| Bulk create/update/delete, submit, cancel, amend, rename, restore, workflows | ffc-bulk-lifecycle |
Reports, search, aggregates, group counts, server methods, raw ffc api | ffc-reports-api |
| Upload/download files, PDFs, comments, assignments, tags, shares, document history | ffc-files-collab |
| Install, sites, login methods, config, doctor, update, cache, desktop app | ffc-setup |
Let an AI client use Frappe through ffc mcp | ffc-mcp |
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

