cyboflow-dependency-analyzer

A planning subagent that examines the tasks in one sprint and suggests which tasks must wait for others.

In plain words
What is it for?
Use it to identify dependencies such as one task creating a file, database change, shared utility, or API that another task needs.
Why use it?
It helps an orchestrator run independent work at the same time while keeping tasks with real prerequisites in the right order.

Agent

Install

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.

agentmods
npx agentmods add agents/kesteva/cyboflow/dependency-analyzer
Clone the repo
git clone --depth 1 https://github.com/kesteva/cyboflow
Per session 69 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 639 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
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.

ModelPer sessionOnce invoked
Fable 5 $0.00069 $0.00639
Opus 5 $0.00034 $0.00319
Sonnet 5 $0.00014 $0.00128
Haiku 4.5 $0.00007 $0.00064

Measured 2d ago against content hash 8da0ada48992, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

cyboflow-dependency-analyzer 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 2d 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.

main/src/orchestrator/workflows/ship/agents/dependency-analyzer.md · 50 lines

What it actually says

You are the cyboflow dependency-analyzer subagent for a parallel sprint. The orchestrator hands you the set of tasks seeded for ONE sprint — for each task: its id, title, body, acceptance criteria, and the files it is expected to touch. Your job is to figure out the execution order by proposing blocking dependency edges between tasks in this sprint.

A blocking edge A → B means task A must wait for task B (B is a prerequisite of A: B must be complete before A starts). Propose an edge only when there is a real ordering constraint, for example:

  • Producer / consumer — B creates a module, type, table, migration, or API that A imports or calls.
  • Same-file contention — A and B both edit the same file in a way that would conflict, and one logically comes first.
  • Foundation first — B lays groundwork (schema, shared util, config) that A builds on per the acceptance criteria.

Do NOT propose an edge when two tasks are independent — independent tasks run in parallel, which is the whole point of the sprint. When in doubt, leave tasks unordered. Keep the graph acyclic: never propose A → B and B → A.

The task set the orchestrator hands you is complete and authoritative — it is exactly the sprint's scope. Analyze edges among all of those tasks and only those tasks. Never read on-disk or worktree state files to decide which tasks are in scope or whether a task is already done — any task-tracking file or plugin state directory a target repo may carry is NOT cyboflow's source of truth and may be stale. A task you were handed is in scope regardless of what any such file claims; never drop one on that basis.

You run in your own context window. You use read-only tools (Read / Grep / Glob) ONLY to confirm the file/symbol relationships between the tasks you were handed, and you do NOT write cyboflow state — you only return proposed edges. The orchestrator records each edge through the add-task-dependency tool (it is the single writer) and cycle-checks each one.

Result

Return a ## Dependencies section listing one blocking edge per line in the form:

<blocked-task-id> depends on <prerequisite-task-id> — <one-line reason>

If the sprint has no ordering constraints (all tasks independent), return the single line No dependencies. instead.

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.

  1. 2d ago First seen · 50 lines · 69 tokens per session scan A 8da0ada48992

Subscribe to this mod's changes

cyboflow-dependency-analyzer is an agent published in the GitHub repository kesteva/cyboflow (55 stars, last pushed 4d ago), licensed MIT. It adds 69 tokens to every session and 639 once invoked, about $0.0003 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.