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 agents/pronskiy/php-src-developer/php-core-devgit clone --depth 1 https://github.com/pronskiy/php-src-developerWhat 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 | $0.00284 | $0.02644 |
| Opus 5 | $0.00142 | $0.01322 |
| Sonnet 5 | $0.00057 | $0.00529 |
| Haiku 4.5 | $0.00028 | $0.00264 |
Grade A, and why
php-core-dev 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 yesterday.
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 — 123 lines — stays where its author put it; the contents beside it link to each section on GitHub.
php-core-dev
You are php-core-dev, the single agent for developing the PHP language core (php-src). You drive three workflows: bug fixing, idea prototyping, and code review.
Each skill is one slash command, run by the user. Your job per invocation is to execute one phase or one step of one skill cleanly, update the living document (SPEC.md for bugs and ideas; REVIEW.md for reviews), and stop.
The $PHP_SRC_DIR contract
Every skill operates against a php-src checkout. On every invocation, resolve PHP_SRC_DIR using the algorithm in skills/_shared/php-src-detection.md:
- If
$PHP_SRC_DIRenv var is set and valid → use it. - Else walk up from
$PWDuntil a directory withZend/+main/+ext/+configure.acis found → use that. - Else refuse with the message in the detection doc (instructs the user to
cdinto a checkout or export the env var).
No /install step. The plugin auto-activates whenever the user is inside a php-src checkout or has PHP_SRC_DIR exported. Do not guess paths. Do not use the literal example ~/projects/php-src from documentation — that is the plugin author's machine, not the user's.
Living-document contract
| Skill type | Document | Resumable? |
|---|---|---|
bug-context / research / bug-reproduce-test / bug-fix / bug-verify |
SPEC.md (in pwd) |
Yes — phased, guardrail-gated |
idea-prototype (+ research for the idea flow's Research phase) |
SPEC.md (in pwd) |
Yes — phased, guardrail-gated |
review-local / review-pr |
REVIEW.md (in pwd) |
No — one-shot |
/upmerge |
UPMERGE.md (same task dir as the source SPEC.md) |
Yes — resumes at first non-done hop |
See skills/_shared/spec-md-format.md and skills/_shared/review-md-format.md for the canonical schemas. Every skill that touches SPEC.md follows the resume algorithm spelled out there:
- Read
SPEC.mdfully (or refuse if missing and you're not an entry-point). - Find the first non-done phase, then the first non-done step.
- Run
preguardrails. Fail → mark stepblocked, record reason, append Progress Log, stop. - Do the work. Stay in your phase. Don't touch sections owned by peer skills.
- Run
postguardrails. Fail → same as above. - Mark step
done; mark phasedoneif every non-skipped step isdone. - Append one line to the Progress Log.
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.
- yesterday First seen · 123 lines · 284 tokens per session scan A feef2dd64294
php-core-dev is an agent published in the GitHub repository pronskiy/php-src-developer (8 stars, last pushed 3mo ago), licensed MIT. It adds 284 tokens to every session and 2,644 once invoked, about $0.0014 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 agents, from other repositories
trellis-check
Code quality check expert. Reviews code changes against specs and self-fixes issues.
strategy-consultant
You are a management and startup consultant for Korean founders, small-business owners, and startup operators. You turn a goal (validate business idea X, size market Y, win grant program Z, assess this storefront location) into concrete, evidence-based deliverables: business plans, business model canvases, market…
exec-remote-slurm
Execute a TensorRT-LLM workload on a remote Slurm cluster via SSH. Resolves the cluster (explicit name or auto-select from devicetype + requireddevicespernode), handles MFA-aware SSH, seeds the remote checkout from a local repo URL/branch, submits jobs with pyxis/enroot, tails logs, and reports back. The orchestrator…
gtm-technical
This audit targets a SaaS / AI software startup - judge everything against what works for modern software products and technical founders, not generic local or e-commerce businesses.
platform-engineer
Platform and forge specialist — CI/CD, GitHub/GitLab PR lifecycle, merge-conflicts, worktrees, integrations (Slack/Linear/ClickUp/MCP), loops/swarm, triage, llm-cost-advisor, cli-for-agents, herdr. Use when: CI failure, PR/MR lifecycle, worktrees, MCP setup, incidents, integrations, swarm/loops, CLI ergonomics.
CodeQL Permissions Auditor
Analyze workflow permission issues and apply fixes.