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 skills add joinwell52-AI/CodeFlowMu-open --skill pm-long-horizon-planninggit clone --depth 1 https://github.com/joinwell52-AI/CodeFlowMu-openWrote 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.
[](https://agentmods.dev/skills/joinwell52-ai/codeflowmu-open/pm-long-horizon-planning)<a href="https://agentmods.dev/skills/joinwell52-ai/codeflowmu-open/pm-long-horizon-planning"><img src="https://agentmods.dev/badge/skills/joinwell52-ai/codeflowmu-open/pm-long-horizon-planning/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.
<a href="https://agentmods.dev/skills/joinwell52-ai/codeflowmu-open/pm-long-horizon-planning"><img src="https://agentmods.dev/badge/skills/joinwell52-ai/codeflowmu-open/pm-long-horizon-planning.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00136 | $0.01764 |
| Opus 5 | $0.00068 | $0.00882 |
| Sonnet 5 | $0.00027 | $0.00353 |
| Haiku 4.5 | $0.00014 | $0.00176 |
Grade A, and why
pm-long-horizon-planning 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 12d 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.
How it starts
The opening of the file, as written. The whole thing — 116 lines — stays where its author put it; the contents beside it link to each section on GitHub.
PM Long-Horizon Planning
Non-negotiable outcome
Compile the source taskbook as governed input. Do not treat it as infallible prose. Produce exactly one current Product Brief, or stop with needs_admin_decision when a blocking source or feasibility finding exists.
Auto-injection is recommendation only. It is never evidence that this skill was executed.
Required resources
Read these files completely at the named phase:
- Before building the IR, read references/planning-model-contract.md and references/taskbook-audit-findings.md.
- Before rendering the artifact, read references/product-brief-contract.md and assets/long-horizon-product-brief-template.md.
- Before submitting or acting on a decision, read references/planning-gate-contract.md.
Workflow
0. Bind identity and target
- Confirm the caller is PM and is in an active Runtime session.
- Resolve one root
task_id, non-emptythread_key, and canonical Brief path. - Accept a child rework only when
parent,planning_target_task_id, thread, and an ADMIN/Planning-Gate rework authorization all bind it to the root. - Reject only the invalid operation when binding fails. Do not terminate the Session or Task.
- Do not write a root Brief from
CHAT-*with an empty thread.
1. Ingest the complete taskbook
- Read from the first byte through EOF, in chunks when necessary.
- Record absolute source identity, version, SHA-256, byte/line counts, read time, ranges read, reference set, and
read_complete=trueonly after EOF. - Resolve file paths, URLs, commits, ports, versions, task IDs, and named authorities.
- Check references read-only. Do not draft a summary Brief before ingestion completes.
2. Audit before planning
- Classify each substantive statement as
normative_constraint,target_design,current_fact,estimate,assumption,reference, orauthority_boundary. - Assign stable
REQ-0001IDs with source lines and modality. - Audit structure, contradictions, references, fact drift, authority, feasibility, roles/acceptance, recovery, reproducibility, and state vocabulary in that order.
- Record findings using the finding contract. Distinguish
blocking,warning, andinfo. - For infeasibility, show a minimal conflicting requirement set and candidate resolutions. Never choose which requirement becomes void.
- Normalize editorial errors only when scope, role, budget, date, gate, acceptance, and authority stay unchanged.
- On any blocking finding, set artifact status to
needs_admin_decision, report the finding, and stop for ADMIN. Do not silently repair intent.
What ships with it
11 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- agents/openai.yaml 307 B
- assets/long-horizon-product-brief-template.md 681 B
- references/planning-gate-contract.md 1.6 KB
- references/planning-model-contract.md 3.3 KB
- references/product-brief-contract.md 2.0 KB
- references/taskbook-audit-findings.md 2.6 KB
- scripts/check_budget_schedule.py 6.2 KB runs code
- scripts/tests/fixtures/source-taskbook.md 74 B
- scripts/tests/test_validators.py 5.0 KB runs code
- scripts/validate_planning_model.py 6.6 KB runs code
- scripts/verify_requirement_coverage.py 4.1 KB runs code
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.
- 12d ago First seen · 116 lines · 136 tokens per session scan A a716d97b080f
pm-long-horizon-planning is a skill published in the GitHub repository joinwell52-AI/CodeFlowMu-open (2 stars, last pushed 18d ago), licensed MIT. It adds 136 tokens to every session and 1,764 once invoked, about $0.0007 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 2026-08-31.
Other skills, from other repositories
issue-creation
Trigger: issue creation, bug reports, feature requests, or issue approval. Create and triage GitHub issues from repository evidence.
sdd-tasks
Break an SDD change into implementation tasks. Trigger: orchestrator launches task planning for a change.
systemic-issue-triage
Trigger: new issue, bug report, triage, backlog, issue flood, community report, root cause, dead-end, blocked user. Attack issues by root class, never one-by-one; fixes must shrink the system, not grow it.
init-workspace-flow
Workflow for initializing or upgrading a workspace: context, discovery, documentation, etc.
arrange-workspace-flow
Workflow for arranging a workspace: layout, reference source code, business/technical context, ecosystem setup.
ijfw-workflow
Use when the user says: 'build', 'create', 'plan', 'new project', 'brainstorm', 'design', 'UI', 'website', 'dashboard', 'app', 'help me build', 'launch', 'book', 'campaign', or anything project-level. Skill body decides Quick vs Deep path.