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 lukasrepublic/agentic-foundry --skill mode-interactivegit clone --depth 1 https://github.com/lukasrepublic/agentic-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/skills/lukasrepublic/agentic-foundry/mode-interactive)<a href="https://agentmods.dev/skills/lukasrepublic/agentic-foundry/mode-interactive"><img src="https://agentmods.dev/badge/skills/lukasrepublic/agentic-foundry/mode-interactive/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/lukasrepublic/agentic-foundry/mode-interactive"><img src="https://agentmods.dev/badge/skills/lukasrepublic/agentic-foundry/mode-interactive.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 | $0.00075 | $0.00661 |
| Opus 5.5 | $0.00030 | $0.00264 |
| Sonnet 5.5 | $0.00015 | $0.00132 |
| Haiku 4.5 | $0.00007 | $0.00066 |
Grade A, and why
mode-interactive 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 1mo 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 — 48 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/foundry:mode-interactive
The default posture. The lean direct loop: ONE accountable context implements inline, runs the live-seam walk, and the operator reviews the merge diff (Regular mode) — the un-forgeable segregation on single-host. No worker dispatch, no queue; the retro shows this beats the dispatch ceremony for most work.
When to trigger
- Default for a single atom / small change. "implement
<atom>interactively", "/foundry:mode-interactive". - Use
mode-autonomousinstead only for scale fan-out over an authorized release.
Procedure
- Confirm the atom is AUTHORIZED (
/foundry:authorizefirst if not). - Implement inline against the frozen
acceptance-contract.yaml(do not weaken checkpoints). - Run the live-seam walk (real boot + Chrome MCP / real seam) against each frozen checkpoint locator.
- Fast lane for breadth (page-context fetch + SSE tail). For many cases against the same seam, a faster faithful method than one-UI-path-at-a-time: drive the app's API from the browser page context (real cookies/session → the real model + DB seam, not a mocked client) and parse the streamed (SSE) response. Use it to exercise breadth quickly.
- Caveat — conversation independence. Independent single-shot page-context calls behave differently from multi-turn/stateful flows; conversation-state independence must be verified, not assumed.
- The UI walk stays authoritative. The fast lane is for breadth only; the UI walk remains authoritative, especially for stateful/multi-turn flows. Do not substitute the fast lane for the authoritative UI walk on those paths.
- Cut the PR (
gh pr create); the worker cuts its PR, the merge floor decides. - Operator reviews the diff + the floor's check results at the merge — the Regular-mode authority. Merge on PASS.
The load-bearing insight
mode-interactive and mode-autonomous differ ONLY at the two ends — who paces the
loop (operator vs /loop) and who reviews before the merge (operator diff-review vs
a fresh-context pr-reviewer pass — advisory, never the merge approval; the floor's green
checks admit either way). The middle (authorized-spec → implement → certification →
the merge floor) is identical and invariant in both.
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.
- 1mo ago First seen · 48 lines · 75 tokens per session scan A 34b1d151dadb
mode-interactive is a skill published in the GitHub repository lukasrepublic/agentic-foundry (1 stars, last pushed 16d ago), licensed MIT. It adds 75 tokens to every session and 661 once invoked, about $0.0003 per session on Opus 5.5. 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 2026-09-03.
Other skills, from other repositories
apex-review
Cross-cutting review of recent work — catches gaps between specialists. Use when asked to "review what we built", "check the work", "pre-launch review", or after completing a significant chunk of work.
spine-review
API and backend code review — REST conventions, auth, validation, error handling, pagination, rate limiting, test coverage. Use when asked to "review this API", "code review", "review backend", or "pre-launch backend check".
bulk-reader
Minimal-tool file-summarization worker (issue #320) dispatched by the read-guard hook to produce a factual summary of an over-threshold file on a cheap, sandboxed worker. Not a general-purpose specialist and never dispatched directly by a human or @techlead — the read-guard hook is its only caller.
g-review
Run the review gate on the current branch diff. Runs the test suite, captures the diff, and dispatches code-lead, which verifies done conditions and reviews the diff itself. Issues MERGE READY or HOLD.
g-help
Context-aware help. With no argument, reads current project state and tells you where you are and what to do next, plus a map of every archive. With a topic or question argument (/g-forge help ), answers it and points you at the right command or archive.
g-refactor
Guided refactor workflow — identify target, pre-analyse, spec, approve, execute, review. Accepts a scope path, an audit milestone file, or runs interactively. Safe-by-default: checks test coverage before execution and runs the full review gate after.