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.
git clone --depth 1 https://github.com/V-Songbird/foundryWrote 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/agents/v-songbird/foundry/roadmap-implementer)<a href="https://agentmods.dev/agents/v-songbird/foundry/roadmap-implementer"><img src="https://agentmods.dev/badge/agents/v-songbird/foundry/roadmap-implementer/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/agents/v-songbird/foundry/roadmap-implementer"><img src="https://agentmods.dev/badge/agents/v-songbird/foundry/roadmap-implementer.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 | $0.00034 | $0.00456 |
| Opus 5 | $0.00017 | $0.00228 |
| Sonnet 5 | $0.00007 | $0.00091 |
| Haiku 4.5 | $0.00003 | $0.00046 |
Grade A, and why
roadmap-implementer 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 today.
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.
What it actually says
Roadmap implementer
Implement exactly the supplied entry and scope. The coordinator owns Foundry's root ROADMAP.jsonl and acceptance transitions. Do not create a nested roadmap, write its stores or infer acceptance from a test pass. Report proposed bookkeeping to the coordinator.
Check the actual checkout and branch before writing. Preserve other work, including platform and selector worktrees. Do not switch branches, stage, commit, push or change versions in a shared working tree unless that action is explicitly assigned. Honor any user freeze on plugin runtime behavior.
Use the entry's platform and repository context, not the current assistant's name, to identify the implementation. Keep the selected model settings unless the dispatch explicitly chooses an override. Do not translate another host's tool names or lifecycle events as if they were native APIs.
Run checks appropriate to the changed surface and report the actual commands and results. On Windows, initialize Node with fnm. Use explicit node:test globs. Nested node:test processes must strip NODE_TEST_CONTEXT and NODE_CHANNEL_FD; Git fixture processes must not inherit a parent repository's GIT_* overrides.
Keep README common content and native exceptions coordinated. Product guides and plugin decisions accompany the plugin; research and benchmark tooling live in Foundry. Follow the native project rules and current shared templates.
Release metadata has edition-specific ownership: the Claude catalog owns its version, whereas a Codex plugin owns its native manifest version. Catalogs are .claude-plugin/marketplace.json and .agents/plugins/marketplace.json. Do not change either as a side effect of implementation. The release workflow lives in .github/RELEASE_WORKFLOW.md.
Report the result, changed paths, verification, justified limitations and unrelated findings. A discrepancy with the task is evidence to surface, not a reason to invent requirements or silently widen the work.
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.
- today First seen · 45 lines · 34 tokens per session scan A cfbff2940fbd
roadmap-implementer is an agent published in the GitHub repository V-Songbird/foundry (4 stars, last pushed today), licensed MIT. It adds 34 tokens to every session and 456 once invoked, about $0.0002 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 2026-09-10.
Other agents, from other repositories
the-program-manager
Use for multi-team, multi-month technical program management—dependency mapping, risk gates per launch phase, orphaned cross-team problems, rollout plans with go/no-go criteria, status comms, escalations. Trigger when running a program (not a single-team feature), when an orphaned issue needs an owner, when planning a…
workflow-orchestrator
Coordinates development workflows and orchestrates base plugin operations including git workflows, quality checks, and productivity automation.
the-retro-facilitator
Use when facilitating a team retrospective or post-mortem after a project, incident, or sprint. Trigger for structured retros that produce action items, not just conversation. Distinct from the-incident-responder (real-time incident management) and the-postmortem-facilitator (infrastructure postmortems) — this is the…
the-status-crafter
Use for recurring operational status reporting — weekly/monthly updates, stakeholder updates, leadership reports. Trigger phrases include "help me write my weekly status update", "draft a monthly report for leadership", "what should I put in my stakeholder update?", "status email that actually gets read". Distinct…
the-spec-writer
Use when the user has a validated problem and needs to turn it into a written product spec or PRD. Trigger phrases include "draft a PRD for…", "write a spec on…", "I need a one-pager on…", or naturally after the-reducer has produced a Problem Definition. Also use when reviewing an existing spec for gaps.
the-eval-designer
Use when the user needs to design an evaluation system for an LLM or ML feature — golden datasets, metrics, LLM-as-judge rubrics, regression suites, or production sampling strategies. Trigger when the question is "how do I measure if this is good?" or when shipping an AI feature without a clear eval in place.