Agentic Workflows API Documentation Lungo

What it does

When your agent uses it

• Updating the lungo workflow API
• Schema/openapi/
• Schema/jsonschemas/ changes
• A prompt references a ticket/issue

Example prompts

• “Use the agentic-workflows-api-documentation-lungo skill to author and maintains the human-facing Agentic Workflows API documentation for the lungo…”
• “/agentic-workflows-api-documentation-lungo”

Requirements

• Python 3

Workflow steps

11 steps, taken from the first numbered list in SKILL.md.

1. 1Title + intro - one-paragraph scope covering catalog, instance lifecycle, streaming, and event schema
2. 2Authoritative sources - table linking in-repo specs and example payloads; note any temporary DTO/catalog gaps pending schema consolidation
3. 3Conventions - identifier URI schemes; path-UUID vs payload-URI rule; authentication; storage model; default port
4. 4Endpoint summary - one table row per operation from OpenAPI (purpose, method and path, response type)
5. 5Catalog API - list/catalog/documentation/chat endpoints declared in OpenAPI
6. 6Workflow details and topology response shapes - workflow GET, topology_only, node/edge field breakdown, example topology from catalog data…
7. 7Workflow-instance lifecycle and state - instantiate, list, get state, delete, each with status codes from OpenAPI
8. 8Internal event ingress - events POST (internal), validation, status codes
9. 9Streaming formats - SSE instance stream and NDJSON pattern chat (implementation notes where OpenAPI only declares success)
10. 10Workflow-instance state JSON Schema - link to published schema; full snapshot vs delta; top-level shape; $defs table; event types; worked…
11. 11Frontend integration checklist - end-to-end flow (selectors, list, graph, instantiate, SSE, reconcile, delete)

What it can do on your machine

Read from SKILL.md and the folder at commit 36f1923. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  Shell commands in SKILL.md call:

  • npm
  • uv

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  No URLs in SKILL.md. Its commands use npm and uv, which can reach the network depending on how they are called.

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names no API keys, tokens, secrets or passwords.

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Agentic Workflows API Documentation Lungo loads about 2.6k tokens when it runs. Until then it costs about 210 tokens; SKILL.md has 1,034 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~210

When it runs · the whole SKILL.md, loaded when a task matches

~2.6k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from agntcy/coffeeAgntcy at commit 36f1923, republished under its Apache-2.0 licence (© agntcy). 1,034 words, ~2,617 tokens.

name

agentic-workflows-api-documentation-lungo

description

Authors and maintains the human-facing Agentic Workflows API documentation for the lungo subproject at coffeeAGNTCY/coffee_agents/lungo/docs/workflow-instance_api.md. OpenAPI under coffeeAGNTCY/coffee_agents/lungo/schema/openapi/ is the HTTP contract source of truth (paths, shapes, response statuses): updated manually or via LLM, never generated from code. Covers the catalog API, workflow-instance lifecycle/state, internal event ingress, SSE and NDJSON streaming, event_v1, and topology/use-case shapes. Use when documenting or updating the lungo workflow API; when schema/openapi/ or schema/jsonschemas/ changes; or when a prompt references a ticket/issue, pull request, or file/folder that changes lungo API contracts. DO NOT TRIGGER AUTOMATICALLY. ASK THE USER IF THE SKILL SHOULD BE USED.

- [ ] 1. Read schema/openapi/ and schema/jsonschemas/ (resolve OpenAPI from its entry point)
- [ ] 2. Enumerate endpoints and response statuses from the resolved spec (do not hard-code)
- [ ] 3. If a new status is needed, update OpenAPI first (see OpenAPI status codes)
- [ ] 4. Write/refresh each section per Document structure from the specs
- [ ] 5. For trigger conditions or streaming details OpenAPI omits, read handler code only as needed; surface drift to the user
- [ ] 6. Use real catalog example values where helpful; link full example payloads rather than inlining
- [ ] 7. Apply the writing conventions
- [ ] 8. Verify links, lint, and OpenAPI unit tests (see Verification)
- [ ] 9. If handlers must align with new statuses or shapes, tell the user to run openapi-to-python-lungo (not this skill)
- [ ] 10. If new contract folders or doc sections appear, update Document structure here; keep AGENTS.md in sync if skill scope changes

cd coffeeAGNTCY/coffee_agents/lungo/docs
for p in $(grep -oE '\]\(\.\.?/[^)]+\)' workflow-instance_api.md | sed -E 's/^\]\(//; s/\)$//'); do
  [ -e "$p" ] && echo "OK   $p" || echo "MISS $p"
done

uv run pytest tests/unit/openapi/ -q

© agntcy, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in .agents/skills/domain-lungo/agentic-workflows-api-documentation-lungo of agntcy/coffeeAgntcy.

Open the folder on GitHubat commit 36f1923