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 agentmods add skills/firatcand/forge/plan-tasknpx skills add firatcand/forge --skill plan-taskgit clone --depth 1 https://github.com/firatcand/forgeWhat 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 | $0.00020 | $0.01273 |
| Opus 5 | $0.00010 | $0.00636 |
| Sonnet 5 | $0.00004 | $0.00255 |
| Haiku 4.5 | $0.00002 | $0.00127 |
Grade A, and why
plan-task scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 88 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/plan-task
Preflight: worktree guard
Task-scoped mutating skill — must run from a worktree, not the main checkout. Contract: skills/_shared/worktree-guard.md.
TOPLEVEL="$(git rev-parse --show-toplevel 2>/dev/null || echo '')"
if [ -z "$TOPLEVEL" ]; then
echo "✗ worktree-guard: not inside a git repository" >&2
exit 1
fi
BRANCH="$(git branch --show-current 2>/dev/null || echo '')"
DEFAULT_BRANCH="$(git symbolic-ref --short refs/remotes/origin/HEAD 2>/dev/null | sed 's@^origin/@@' || echo 'main')"
if [ -f "$TOPLEVEL/.forge/worktree-task.json" ]; then
echo "✓ worktree-guard: $TOPLEVEL ($BRANCH)"
elif [ "$BRANCH" = "${DEFAULT_BRANCH:-main}" ]; then
echo "✗ worktree-guard: refusing to run on default branch '$BRANCH' from $TOPLEVEL" >&2
echo " Parallel Claude Code sessions on the main checkout clobber each other's HEAD." >&2
echo " → cd into an existing worktree, or run /pickup-task to create one." >&2
exit 1
else
echo "⚠ worktree-guard: no .forge/worktree-task.json at $TOPLEVEL (branch: $BRANCH)" >&2
echo " Proceeding, but this worktree may be manually-created or stale." >&2
fi
Steps
- Read Linear issue (or phases.yaml task) for current branch
- Determine task type → delegate to relevant specialist subagent (frontend-dev, backend-dev, db-architect, etc.)
- Specialist enters Plan mode: research codebase, propose approach
- Detect architectural forks. Before drafting the plan, enumerate every decision the implementation will require. For each one, classify it (see Fork severity filter below). Anything that passes the filter is a question — not a silent decision.
- Ask per fork, not in bulk. For each qualifying fork, emit an
AskUserQuestioncall followingskills/_shared/question-format.md— re-ground, simplify, recommend, options with concrete trade-offs. Batch up to 4 questions perAskUserQuestioncall; iterate as new forks surface during deeper research. Do not bundle 11 forks into a "Questions decided" table for the user to rubber-stamp. - Write structured plan to
plans/tasks/{LINEAR-ID}.plan.md:- Files to change (predicted)
- Component tree / data flow
- State / data flow
- Edge cases
- Test strategy
- Questions asked & answers applied (not "decisions decided" — list each question, who answered, and the resulting branch in the plan)
- Show plan to user, wait for approval
- On approval: commit plan, unlock /implement
- Emit the second-opinion suggestion hint (one printed line; user types or skips):
Printsforge second-opinion suggest plan-task💡 Suggested next: /second-opinion review-plan (run with FORGE_AUTO_SECOND_OPINION=0 to disable). Silent whenFORGE_AUTO_SECOND_OPINION=0is set orsecond_opinion.auto_enabled: falsein.forge/settings.yaml.
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.
- 2d ago First seen · 88 lines · 20 tokens per session scan A 80f44f1b2543
plan-task is a skill published in the GitHub repository firatcand/forge (13 stars, last pushed 1mo ago), licensed MIT. It adds 20 tokens to every session and 1,273 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 2026-08-30.
Other skills, from other repositories
ospec-goal
Create or advance a full OSpec goal using the current document, task graph, worker, review, and evidence workflow.
ospec
Document-driven OSpec workflow for initialization, change/goal routing, validation, archiving, and durable project knowledge.
ospec-change
Create or advance a lightweight OSpec change using the classic fast workflow.
prospec-verify
Verify Implementation - Run 5+1 dimension audit (tasks, spec compliance, constitution, knowledge-implementation consistency, tests, design consistency) and assign quality grade (S/A/B/C/D). Triggers: verify, audit, quality check, 驗證, 稽核, 品質檢查, 評級.
prospec-archive
Archive Changes - Archive completed changes, generate summary, sync requirements to feature specs, and gate archiving on Knowledge sync. Triggers: archive, spec sync, finalize change, 封存, 歸檔, 收尾, 規格同步.
prospec-knowledge-generate
Generate AI Knowledge - Read raw-scan.md, analyze project structure, autonomously decide module boundaries, and produce Recipe-First module READMEs and index. Triggers: generate knowledge, analyze project, module split, 產生知識, 知識庫, 分析專案, 模組拆分.