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 commands/griddynamics/rosetta/testgen-flow-project-config-loadinggit clone --depth 1 https://github.com/griddynamics/rosettaWrote 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/commands/griddynamics/rosetta/testgen-flow-project-config-loading)<a href="https://agentmods.dev/commands/griddynamics/rosetta/testgen-flow-project-config-loading"><img src="https://agentmods.dev/badge/commands/griddynamics/rosetta/testgen-flow-project-config-loading.svg" alt="Measured on agentmods" 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 | $0.00017 | $0.02652 |
| Opus 5 | $0.00009 | $0.01326 |
| Sonnet 5 | $0.00003 | $0.00530 |
| Haiku 4.5 | $0.00002 | $0.00265 |
Grade A, and why
testgen-flow-project-config-loading 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.
How it starts
The opening of the file, as written. The whole thing — 190 lines — stays where its author put it; the contents beside it link to each section on GitHub.
<testgen_flow_project_config_loading>
<description_and_purpose> Find or create the project config file, obtain project-specific data retrieval configuration from user, and initialize the ticket output directory. </description_and_purpose>
<workflow_context>
- Phase 0 of 7 in
testgen-flow - Input: user request with an Issue Tracker ticket key/URL (Jira
PROJ-123is the canonical example) - Output:
initial-data.md, project config file, initialized state - Prerequisite: user provided an Issue Tracker ticket key or URL </workflow_context>
<phase_steps>
- Parse initial user input for ticket key
- Setup output directory
- Load or create project config
- Obtain project info from user (if config is new)
- Create initial-data file
- Update state file </phase_steps>
<parse_input step="0.1">
- Extract the Issue Tracker ticket key or URL from the user prompt (REQUIRED) — Jira shapes are the canonical examples; adapt parsing to the resolved tracker.
- Accept formats:
- Ticket only:
PROJ-123, a full ticket URL, orAnalyze requirements for PROJ-123. - Ticket + Wiki: the ticket key/URL plus one or more Wiki page URLs, e.g.
Analyze PROJ-123, docs: https://<wiki>/pages/123. - Ticket + other documentation URLs (e.g. Google Drive): the ticket plus any documentation links the project config lists as data sources.
- Ticket only:
- Any non-ticket links pasted here are captured verbatim into
initial-data.mdand carried into Phase 1 data collection, resolved there per the project config's data sources. A recognizable provider URL is valid provider evidence. </parse_input>
<setup_directory step="0.2">
- Create
plans/testgen-{TICKET-KEY}/(full per-ticket layout in<output_directory>below). - Initialize
testgen-state.mdfrom<state_file_template>below. At init the Phase 0 row is[ ] Phase 0: Project Config Loading - In progress(all others Not started); step 0.6 flips it to[x] ... Completedonly after the config and initial-data files exist. </setup_directory>
<load_project_config step="0.3">
- Search for
testgen-project-config.mdatplans/testgen-{TICKET-KEY}/testgen-project-config.md(per-ticket, inside this ticket's feature plan folder — one copy per run, not a shared project-wide file, so parallel sessions and commits never collide). - Branches (exhaustive):
- File exists AND non-empty: skip to step 0.5.
- File missing OR exists but empty: proceed to step 0.4 (do NOT create an empty placeholder file — step 0.4 will write the populated file). </load_project_config>
<obtain_project_info step="0.4">
Contiguous 1–5 sequence. The <example_format_of_question> block below is the verbatim question text used by step 2 — it is not a numbered step and the sequence does not restart after it.
-
Pre-fill from
gain.json(merge evidence; do not force one source): read repository-rootgain.json; usesdlc.issue_tracker(_project),sdlc.wiki(_project), andsdlc.test_management(_project)when populated, plus recognizable provider URLs from the prompt. Explicit user input wins for this run; a missinggain.jsonnever blocks the phase. Name the prefilled providers in the question below so the user confirms rather than re-supplies them. -
USE SKILL
questioning. Ask the user about knowledge base and data retrieval setup using the question text in<example_format_of_question>below (adapt its canonical provider names to the prefilled ones). -
Process the user's answer — confirm the default scheme OR capture their customization.
-
Validate the answer provides sufficient information. Minimum required fields:
- Data sources (which of: Issue Tracker, Wiki, attached docs, other URLs — resolved providers recorded by name)
- Retrieval method per source (MCP-based / direct URL / search-by-keywords)
- Auth assumptions (MCP already configured / token in env / requires per-call OAuth)
Validation failure paths:
- If user said YES to default but the default cannot run in the environment (no MCP, no auth): re-prompt up to 2 times naming exactly which field(s) are absent. After 2 unsuccessful re-prompts, stop Phase 0, record
Phase 0 blocked: default unrunnable — missing <field>intestgen-state.md, and ask the user to supply a complete answer before continuing. - If user said NO but did not name source / method / auth: re-prompt up to 2 times naming the missing fields explicitly. After 2 unsuccessful re-prompts, stop Phase 0, record
Phase 0 blocked: incomplete config answerintestgen-state.md, and ask the user to supply a complete answer before continuing.
-
Apply the
<safety_boundaries>redaction-at-intake gate, then save the validated configuration toplans/testgen-{TICKET-KEY}/testgen-project-config.md(path per step 0.3).
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 · 190 lines · 17 tokens per session scan A 618cdba6bb65
testgen-flow-project-config-loading is a command published in the GitHub repository griddynamics/rosetta (342 stars, last pushed today), licensed Apache-2.0. It adds 17 tokens to every session and 2,652 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 2026-09-03.
Other commands, from other repositories
resolve-conflicts
Resolve merge conflicts on a GitHub PR by merging the base branch into the PR branch in an isolated git worktree, with JSON/markdown-aware conflict resolution.
xpoz-best-practices
Reference docs for Xpoz query syntax, pagination, field selection, platform tools, and authentication. Usage: /xpoz-best-practices.
align
Verify spec-vs-spec alignment and brevity across the intent tree. This is NOT QA — QA checks code against a spec; alignment checks whether specs are consistent with their ancestors and internally tight against the Brevity Contract. The aligner runs three passes per walk.
deploy-check
Pre-flight verification before pushing to main. Reproduces the CI gates locally so the push doesn't fail in production.
validate-fix
Validate a proposed React fix with React-Sentinel assertions or replay checks.
rclone_mkdir
Make the path if it doesn't already exist.